Project Report No. 432

July 2008 Price: £6.00

Understanding the basis of resistance to Fusarium head blight in UK winter wheat

(REFAM)

by

P. Nicholson1, R. Bayles2 and P. Jennings3

1John Innes Centre, Norwich Research Park, Colney, Norwich NR4 7UH 2NIAB, Huntingdon Road, Cambridge CB3 0LE

3Central Science Laboratory, Sand Hutton, York, YO41 1LZ

This is the final report of a Defra-sponsored Link Project (LK0932) that ran for four years from April 2003. The project was sponsored by Defra (£238,068) and HGCA (£225,579, Project No. 2726) with industrial support of £2,000 each from Advanta Seeds UK, Elsoms Seeds Ltd, RAGT Ltd (previously Monsanto UK Ltd) and Nickerson UK Ltd. Further in-kind contributions of £66,222 from industry (Advanta Seeds UK, £17,047, Elsoms Seeds Ltd, £11,325, RAGT Ltd, £17,250, Nickerson UK Ltd, £9,600 and RHM Technology Ltd. £11,000) made the total funding £537,869.

HGCA has provided funding for this project but has not conducted the research or written this report. While the authors have worked on the best information available to them, neither HGCA nor the authors shall in any event be liable for any loss, damage or injury howsoever suffered directly or indirectly in relation to the report or the research on which it is based.

Reference herein to trade names and proprietary products without stating that they are protected does not imply that they may be regarded as unprotected and thus free for general use. No endorsement of named products is intended nor is it any criticism implied of other alternative, but unnamed, products.

1

CONTENTS

Page number ABSTRACT 3 PROJECT SUMMARY 4 FULL PROJECT REPORT 14 1 Introduction 14 2 Materials and methods 17

2.1 Plant materials 17 2.2 Genotyping 17 2.3 Phenotyping for FHB resistance 17 2.4 DON analysis, DNA extraction and quantitative PCR 18 2.5 SSR haplotyping 19 2.6 Statistical analysis 20

3 Evaluation of UK released varieties and selected European winter wheat lines with known FHB resistance 22

3.1 Results 23 3.1.1 Fusarium head blight resistance 23 3.1.2 Relationship between FHB symptoms and the deoxynivalenol (DON) and fungal DNA content of grain 27 3.1.3 Haplotyping with known SSRs linked to FHB resistance QTL 27

3.2 Discussion 33 3.2.1 FHB status of UK winter wheat varieties 33

4 Identification and mapping of QTL for FHB resistance and other agronomic traits in different mapping populations 38 4.1 Spark/Rialto DH population 38

4.1.1 Results 38 4.1.1.1 Trait analysis in the population 38

4.1.1.2 Association between height and disease 39 4.1.1.3 Identification of FHB resistance QTL 39 4.1.1.4 Plant height QTL 41

4.2 Soissons/Orvantis DH population 43 4.2.1 Results 43

4.2.1.1 Fusarium head blight resistance 43 4.2.1.2 Correlations, ANOVA and broad sense heritability 43 4.2.1.3 FHB and Plant height QTL 44 4.2.1.4 Phenotyping of Rht-B1b and Rht-D1b near-isogenic lines 44

4.3 RL4137/Timgalen Recombinant Inbred Population 50 4.3.1 Results 50

4.3.1.1 Performance of RILs 50 4.3.1.2 Mapping of QTL for FHB resistance and associated traits 51

4.4 Discussion 54 4.4.1 FHB resistance QTL in Spark and the role of Rht-D1b

semi-dwarfing allele in FHB susceptibility 54

2

4.4.2 FHB resistance QTL identified in Soissons and the role of Rht-B1b and Rht-D1b semi-dwarfing alleles in resistance 58 4.4.3 FHB resistance QTL detected in RL4137/Timgalen cross 60

4.4.3.1 Disease assessment and QTL mapping for FHB resistance 60 4.4.3.2 QTL detection for other traits and their role in FHB resistance 61

5 Identification of new sources of resistance from CIMMYT wheat lines 65

5.1 Background 65 5.2 Results 66

5.2.1 Characterisation of FHB resistance in wheat varieties inoculated with toxin producing and non-producing species in the glasshouse 66

5.2.1.1 Spray inoculation 66 5.2.1.2 Point inoculation 66

5.2.2 Characterisation of resistance 67 5.2.3 Spray and point inoculation of Mercia winter wheat with DON and NIV producing isolates of F. graminearum 68 5.2.4 Spray and point inoculation of CIMMYT wheat lines with NIV and DON-producing isolates of F. graminearum 69

5.3 Discussion 75 6 Evaluation of FHB resistance within UK spring barley varieties 78 6.1 Background 78 6.2 Materials and Methods 78 6.3 Results 79 6.4 Discussion 81 7 Acknowledgements 82 8 References 83

3

ABSTRACT

Fusarium head blight (FHB) of wheat is caused predominantly by Fusarium graminearum and F. culmorum although other Fusarium species and Microdochium majus and M. nivale are also important in some regions. The disease can often contaminate the grain with mycotoxins such as deoxynivalenol (DON) and nivalenol (NIV). Of 53 UK National List varieties tested for response to FHB, only three (Soissons, Spark and Vector) had significant stable resistance over trials. UK barley varieties differed significantly in FHB resistance. Three FHB resistant varieties were studied to identify the location of quantitative trait loci (QTL) associated with FHB resistance. Analysis of Spark and Soissons was combined with study of near-isogenic semi-dwarf lines for Rht1 and Rht2. Our results demonstrated that Rht2 is associated with a significant increase in susceptibility to initial infection (Type I resistance) while being largely unaffected in resistance to spread within the spike (Type II resistance). In contrast, Rht1 conferred no negative effect on FHB resistance, even conferring a very minor positive effect in one trial. Under high disease pressure both Rht1 and Rht2 significantly decreased Type 1 resistance. However, while Rht2 had no effect on Type 2 resistance Rht1 significantly increased Type 2 resistance. Enhanced susceptibility associated with Rht2 is probably due to linkage to deleterious genes rather than to pleiotropy and the positive effect of Rht1 on FHB resistance is due either to pleiotropy conferring Type 2 resistance or very tight linkage to resistance genes. In the third variety (RL4137), we identified FHB resistance QTL on chromosomes 1B and 2B. Correlation for resistance to F. culmorum (DON-producer) and M. majus (non toxin-producer) was moderate across 29 European varieties following spray inoculation. Following point inoculation M. majus was not able to spread. Type 2 resistance appears to be important to restrict spread of DON-producing isolates of some species but may be largely irrelevant for other pathogens. Spread of a NIV-producing isolate of F. graminearum was much slower than that of a DON producing isolate. These isolates were used to identify and characterise new sources of FHB resistance among 300 lines from CIMMYT. 60 lines were shown to have moderate/high levels of FHB resistance. A few lines possessed a high level of Type I resistance only whereas a greater number possessed both Type I and Type II resistance. These lines merit further study as potential sources of novel FHB resistance. Furthermore, we propose that spray inoculation with an appropriate aggressive non DON-producing FHB pathogens may be used to identify the Type I FHB resistance component in wheat.

4

PROJECT SUMMARY

Introduction Fusarium head blight (FHB) of wheat (also known as Fusarium ear blight), is caused

by several fungal species that produce similar symptoms. Fusarium graminearum is

the major pathogen worldwide, while F. culmorum tends to predominate in maritime

regions. Fusarium avenaceum and F. poae are also frequently associated with FHB,

particularly in Northern Europe. In addition to the true Fusarium species, two

Microdochium species, M. majus and M. nivale, also cause FHB and are particularly

prevalent where cooler, wetter conditions prevail such as in the UK.

FHB is of particular concern because many of the Fusarium species produce

mycotoxins in infected grain and pose a risk to human and animal consumers. The

most common mycotoxin in blighted grain is the trichothecene deoxynivalenol (DON),

produced by Fusarium graminearum and F. culmorum. A second, closely related

trichothecene, produced by certain isolates of these species is nivalenol (NIV).

The development and deployment of FHB-resistant cultivars is generally

accepted as the most cost-effective and environmentally benign way to minimise

disease and potential risk to consumers. However, resistance to FHB is quantitatively

inherited and the influence of environment on disease makes reliable disease

screening (phenotyping) difficult. To date no variety has been found to be immune to

FHB. Advances in phenotyping, combined with statistical methods to detect regions on

wheat chromosomes harbouring genes for resistance to FHB, have led to the

identification of numerous so-called quantitative trait loci (QTL) for FHB resistance.

Several important sources of FHB resistance have been identified among Chinese

spring wheat varieties that have been deployed in breeding programs worldwide. The

resistance of some of these varieties, notably Sumai 3, appears to be controlled by a

few genes of major effect and hence may be amenable for use in marker-assisted

breeding programmes that can greatly facilitate the rapid introgression of resistance

into new varieties.

Although a number of European winter wheat varieties show resistance to FHB,

in only a few varieties has the basis for the resistance been genetically characterised.

In contrast to the resistance in some spring wheat varieties, FHB resistance in winter

wheat germplasm appears to be due to numerous QTL of moderate to small effect.

DNA markers, such as ‘simple sequence repeats’ (SSRs) can be used to infer the

origin and genetic relationship between FHB resistance QTL. This information is used

5

by plant breeders to identify appropriate parental combinations in order to ‘pyramid’

resistance genes when developing new varieties.

Due to the previously low incidence of FHB in the UK, a comprehensive

assessment of resistance in elite UK winter wheat germplasm has not yet been

undertaken. One aim of the current study was to assess the FHB reaction of winter

wheat and spring barley varieties currently on the UK National List (2003). A second

aim was to compare the sizes of SSR markers at FHB QTL in Chinese varieties with

those in UK and European winter wheat varieties to establish whether different genes

might be responsible. Diversification of resistance sources in breeding programs

should reduce the risk of the emergence of virulent pathogen strains.

An important aim of this project was to understand the genetic basis of FHB

resistance in selected wheat varieties. Initial studies showed Soissons and Spark to be

the most resistant UK varieties and QTL analysis was undertaken of Spark x Rialto and

Soissons x Orvantis doubled haploid populations. A third population consisted of

recombinant inbred lines from RL4137 (FHB resistant) x Timgalen. Field and

polytunnel disease trials were established and each line scored at several locations.

Genetic maps were produced for all three crosses and QTL analysis carried out for FHB

resistance traits. Several authors have reported a negative relationship between plant

height and FHB resistance and, for this reason we also undertook QTL analysis for

selected morphological traits to determine the genetic basis of FHB resistance and

identify any associations between resistance to FHB and characteristics such as plant

height.

The mycotoxin DON is required to facilitate the spread of fungus from the point

of infection into other parts of the wheat head via the rachis. Comparison of DON and

NIV chemotypes of F. graminearum suggest that DON production is associated with

greater disease causing potential. The reduced aggressiveness of NIV, relative to DON

producing isolates (chemotypes), may stem from the much lower phytotoxicity of NIV

towards wheat. In contrast to F. graminearum and F. culmorum, Microdochium majus

is not known to produce mycotoxins.

Resistance of wheat to FHB appears to be horizontal and non-species specific

with no clear evidence for any differential effect on different pathogen species. An

additional aim of the current study was to investigate whether FHB resistance

effective against toxin-producing species is also effective against non toxigenic

species. Two components of host resistance to FHB are widely recognised: resistance

to initial infection (Type I resistance) and resistance to spread within the spike (Type

II). It is generally accepted that single spikelet (point) inoculation assesses Type II

6

resistance only while spraying a conidial suspension on spikes and scoring disease

incidence on a plot basis assesses Type I resistance. However, accurate assessment of

Type I resistance can be hindered by differences among varieties in their degree of

Type II resistance leading to altered disease severity.

Several major QTL conditioning Type II resistance have been reported but only

a few studies have identified QTL for Type I resistance. This may reflect a paucity of

Type I resistance, but it is also probable that the need to infer Type I resistance is

hampering the identification of this form of resistance. If species, or isolates, that

produce little or no toxin can infect but not spread within the spike they might be used

as tools to identify Type I resistance. We undertook studies to establish whether

particular species, or isolates, of FHB causing pathogens could be used to facilitate the

identification of Type I resistance and established trials to identify and characterise

potential sources of FHB resistance among a collection of wheat lines obtained from

the International Maize and Wheat Improvement Center (CIMMYT), Mexico.

Materials and methods

Plant materials, map construction and QTL analysis

The Fusarium head blight (FHB) reaction of 53 varieties from the (2003) National List

of winter wheat varieties approved for sale in the United Kingdom (UK) was compared

with 19 reference cultivars from Continental Europe which had previously been

characterised for resistance by collaborative partners. A similar assessment was made

for UK spring barley varieties.

Recombinant inbred lines (RILs) of a cross between RL4137 and Timgalen and

doubled haploid (DH) populations: Spark/Rialto, Soissons/Orvantis, was used in the

study. Near isogenic lines (NILs) differing in their Rht1 and Rht2 semi-dwarfing alleles

in Mercia and Maris Huntsman background were used to assess the relationship

between FHB resistance and height.

Thirty winter wheat varieties were used to assess the efficacy of FHB resistance

against toxin-producing and non-producing species. The level and type of FHB

resistance was assessed in a collection of 300 lines from a CIMMYT FHB resistance

breeding programme.

FHB disease screening was carried out in field, glasshouse and polytunnel

experiments conducted either at the John Innes Centre (JIC), Norwich, National

Institute for Agricultural Botany (NIAB), Cambridge, Central Science Laboratory (CSL),

York or at sites of participating commercial partners (Nickerson Seeds Ltd, Advanta

Seeds (UK) Ltd. Elsoms Seeds Ltd.). Plants were inoculated at mid-anthesis with

7

conidial suspensions of highly virulent DON producing isolates of Fusarium culmorum,

F. graminearum or a NIV producing isolate of F. graminearum or Microdochium majus

(non toxin-producer), either by spray or point inoculation as appropriate. Following

spray inoculation, disease was assessed several times and expressed as area under

the disease progress curve (AUDPC). For point inoculation experiments disease was

generally measured as the number of spikelets showing symptoms. In addition to

measuring disease, selected morphological traits such as presence/absence of awns,

plant height (PH) and weight of infected spikelets (WIS) were recorded.

The genetic maps for the mapping populations were constructed using SSRs,

Amplified Fragment Length Polymorphism (AFLP) and Diversity Array Technology

(DArT) markers. Linkage maps were constructed and QTL detection was carried out by

Interval Mapping (IM) and using the Multiple QTL Model (MQM). The QTL that

explained more than 10 % of the variance (R2) in at least one

environment/experiment were classified as major QTL and those explaining less than

10 % as minor QTL.

DON analysis, DNA extraction and quantitative PCR

The DON content of milled grain was assessed using an enzyme linked immuno-assay

(ELISA) according to the manufacturer’s instructions. DNA extraction and competitive

PCR were performed using specific primers developed within our laboratory and the

amount of fungal DNA was expressed as a percentage of the total DNA content of the

sample.

SSR haplotyping

Comparison of SSR allele sizes from known, genetically characterised FHB resistant

and susceptible cultivars from Asia, Europe and USA were used to infer the origin of

FHB resistance QTL in the trial varieties and to identify potentially novel loci.

Statistical analyses

All the statistical analysis was performed using GenStat for Windows 9th edition

(copyright Lawes Agricultural Trust, Rothamsted Experimental Station, UK).

8

Results

Fusarium head blight status of UK winter wheat and spring barley

varieties

The Fusarium head blight (FHB) reaction of 53 varieties from the (2003) National List

of winter wheat varieties was compared with 19 reference cultivars from Continental

Europe that had previously been characterised for resistance. Of the National List

varieties tested, only Soissons, Spark and Vector had stable resistance over trial sites.

In addition, under moderate disease pressure, a total of 24 National List varieties had

levels of the trichothecene mycotoxin deoxynivalenol (DON) above the EU limit of 1.25

parts per million (ppm) in grain. Significant and consistent differences in resistance to

FHB were found among the barley varieties and disease levels were found to correlate

with DON content.

Comparison of SSR allele size was used to infer the origin of FHB resistance and

to identify germplasm with potentially novel loci. A total of 17 SSR loci were selected

from published studies of resistance on chromosomes 3BS, 5A and 6B associated with

resistance in the Chinese cultivar Sumai-3, chromosomes 1B, and 5A associated with

resistance in the Romanian cultivar Fundulea F201R (F201R) and chromosome 5AL

associated with resistance in the French cultivar Renan. No variety appeared to

possess FHB QTL similar to those of Sumai-3 (3BS, 5A and 6B), F201R (1B) and

Renan (5A). However, the highly resistant German reference cultivar Petrus had an

identical haplotype to F201R on 1B indicating that this cultivar has an allelic QTL at

that location.

FHB resistance quantitative trait loci (QTL) detected in Spark x Rialto Spark is more resistant to FHB than most other UK winter wheat varieties but the

genetic basis for this is not known. Spark carries no known mutation at either Rht-B1

or Rht-D1 loci whereas Rialto carries the Rht2 allele at the Rht-D1 locus. A mapping

population from a cross between Spark and the FHB susceptible variety Rialto was

used to identify QTL associated with resistance. QTL analysis across environments

revealed nine QTL for FHB resistance and four QTL for plant height. Spark contributed

seven QTL (2A, 3A, 4D (2 QTL), 5A, 6A, 7A) while two QTL were derived from Rialto

(1B, 3A). Two QTL for PH were contributed by Spark (4D, 6A) and two by Rialto (2A,

3B). One FHB QTL was coincident with the Rht-1D (Rht-2) locus and accounted for up

to 51% of the phenotypic variance. None of the other height QTL was associated with

FHB resistance.

9

FHB resistance QTL detected in Soissons x Orvantis Soissons is one of the most resistant varieties grown in the UK. Soissons carries Rht-

D1a (Rht1) while Orvantis carries (Rht2). QTL analysis of FHB revealed only a single

major FHB QTL on chromosome 4D effective in all three field trials. Soissons (Rht-

D1a) contributed the FHB resistance allele. Although this QTL is in the region of the

Rht-D1 locus, the peak of the QTL was closer to an adjacent marker. The major PH-

QTL associated with the Rht-B1 (Rht1) locus also co-localised with a putative minor

FHB QTL on 4BS but surprisingly, and in contrast to the effect around the Rht-D1

locus, FHB resistance was associated with the Rht-B1b (Rht1) allele (contributed by

Soissons) rather than the Rht-B1a (tall) allele (Orvantis) as might have been expected

if the effect were due to differences in plant height. Putative QTL for FHB resistance,

often appearing in more than one trial, were also detected (1BL, 3BL, 4BS, and 7AL).

Soissons contributed all the alleles for FHB resistance except that on 1B.

Association between FHB susceptibility and plant height determined by

Rht alleles

In both the Spark x Rialto and Soissons x Orvantis populations a major FHB QTL was

found on chromosome 4D at, or close to, the Rht-D1 locus with the Rht2 allele

associated with FHB susceptibility. In the Soissons x Orvantis population a minor FHB

QTL was detected on 4BS at the Rht-B1 locus but surprisingly, and in contrast to the

effect around the Rht-D1 locus, FHB resistance was associated with the Rht-B1b allele

(Rht1), contributed by Soissons, rather than the Rht-B1a allele (Orvantis) as might

have been expected if the effect were due to differences in plant height. These results

indicated that the Rht-1 and Rht2 differ in their effects on FHB resistance. The

influence of the Rht-B1b and Rht-D1b alleles on FHB resistance was further

investigated using both Mercia and Maris Huntsman near isogenic lines. Under high

disease pressure both Rht-B1b and Rht-D1b significantly decreased Type I resistance

(resistance to initial infection). However, Rht-D1b had no effect on Type II resistance

(resistance to spread of the fungus within the spike), while Rht-B1b significantly

increased Type II resistance.

Mapping of QTL associated with Fusarium head blight in RL4137

RL4137 is a FHB resistant line derived from the Brazilian variety Frontana. The study

used 90 recombinant inbred lines (RIL) derived from a cross between RL4137 and the

moderately FHB resistant variety Timgalen. QTL analyses identified a total of six FHB

resistance QTL (1B, 2B, 3A, 6A, 6B and 7A). In all but one instance, the alleles from

10

RL4137 had a positive effect on FHB resistance. The FHB QTL on 1B, 2B and 6B were

detected in multiple trials, with alleles from RL4137 contributing a positive QTL for

resistance on 1B and 2B and the alleles from Timgalen contributing a positive QTL for

resistance on 6B. Our study also identified three QTL for plant height (PH) (2B, 4A and

5B), two QTL for weight of infected spikelet (WIS) from infected ears (2B and 6A) and

one QTL for awns (2B). The QTL mapped on 2B for PH, WIS and awns co-localized

with that for FHB resistance.

Assessing non-specificity of FHB resistance, development of

methodologies to detect type I resistance and identification of novel

sources of FHB resistance.

Fusarium head blight (FHB) of wheat is caused predominantly by Fusarium

graminearum and F. culmorum although other Fusarium species and Microdochium

majus and M. nivale are also important in some regions. The few reports to date

suggest that FHB resistance is effective against all pathogen species. However,

because biosynthesis of DON has been shown to be critical for spread of F.

graminearum within the spike we reasoned that isolates that do not produce DON

might be unable to spread within the spike. Appropriate non DON-producing isolates

might be used to reveal Type I resistance (resistance to initial infection) without the

confounding effects of differences in Type II resistance (resistance to spread within

the spike). In initial experiments, thirty European winter wheat varieties were spray

and point inoculated with a DON-producing isolate of F. culmorum or an isolate of M.

majus in glasshouse tests. Resistance to the two pathogens following spray

inoculation was well correlated whereas, following point inoculation, no correlation

was observed because M. majus was unable to spread beyond the inoculated spikelet.

However, spray inoculation with M. majus produced only low levels of disease making

it unsuitable for use in routine screening. In a second set of experiments we found

that a NIV-producing isolate of F. graminearum caused high levels of disease following

spray inoculation but spread only very slowly within the spike. Comparative spray and

point trials using DON and NIV-producing isolates of F. graminearum were carried out

to characterise a set of wheat lines for their Type I and II FHB resistance components.

From 300 lines, three possessed a high level of Type I resistance and several

possessed high levels of both Type I and Type II resistance.

11

Discussion

FHB resistance of UK National List entries was assessed and compared to resistant

European cultivars in three contrasting environments. Although the resistance of

varieties correlated well across sites, some varieties differed markedly in resistance

between sites. This effect is due primarily to the differences in environment at the

three sites but we also found evidence that suggests that the period of optimal

susceptibility to FHB can differ significantly between cultivars. Assessment of visual

disease, DON and FDNA data from three inoculation timings indicated that some

varieties had a very narrow (three days or less) period of optimal susceptibility

whereas others remained highly susceptible over a six day period. These findings

make it imperative that lines are inoculated at the same developmental stage when

carrying out trials at different locations or in different years.

Among the National List varieties (2003), only Soissons, Spark and Vector

showed evidence for moderate FHB resistance. Even under the moderate disease

pressure 24 National List varieties had DON levels which were above the proposed EU

action limits of 1.25 parts per million (ppm). These results indicate that a significant

effort will be required by the UK plant breeding community to improve overall levels of

FHB resistance. For barley, significant genetic variation exists for resistance to FHB

among UK varieties. In general, decreased symptoms correlated with reduced DON

content of grain for both wheat and barley.

None of the FHB resistant varieties from the UK or mainland Europe had SSR

haplotypes indicating that their resistance is derived from Sumai-3. Thus the

introduction of potent FHB QTL from this source should complement those of the FHB

resistant European varieties to increase overall levels of resistance. While several

varieties carried the 1RS-1BL rye translocation that confers type II resistance, only

the resistant German variety Petrus carries the entire region associated with this QTL.

Selection of lines that carry all the markers relating to this QTL should ensure a

minimal level of FHB resistance among varieties carrying the 1RS-1BL translocation.

The genetic basis of FHB resistance of Soissons and Spark, the two most resistant UK

varieties, was assessed along with that of a Frontana-derived line. QTL analysis of the

Spark x Rialto population revealed that the main effect (up to 51% of phenotypic

variance) was coincident with the dwarfing locus Rht-1D (Rht2). No other height QTL

was associated with FHB resistance in this cross. Surprisingly, the main effect in the

Soissons x Orvantis population also occurred close to the same locus. However, in this

population, the QTL peak lay over an adjacent marker suggesting that the

susceptibility is due to a tightly linked gene rather than pleiotropy associated with

12

Rht2. In contrast, Rht1 (carried by Soissons) conferred no negative effect on FHB

resistance, even conferring a very minor positive effect in one trial. Additional

experiments with near-isogenic lines supported these findings. Under high disease

pressure both Rht-B1b and Rht-D1b significantly decreased resistance to initial

infection. However, while Rht-D1b had no effect on resistance to spread within the

spike, Rht-B1b significantly increased resistance to spread. Combined with the

evidence from the population studies above our study suggests that the enhanced

susceptibility of Rht-D1b allele is due to linkage to deleterious genes rather than to

pleiotropy and that the positive effect of Rht-B1b allele on FHB resistance is due either

to pleiotropy conferring Type II resistance or very tight linkage to resistance genes.

The majority of UK winter wheat varieties are highly susceptible to FHB and

almost all these carry the semi-dwarfing Rht-D1b (Rht2) allele. Neither Soissons nor

Spark carry Rht-D1b: Soissons possesses Rht-B1b (Rht1) and Spark has a tall (rht)

genotype with its reduced height being due to non-Rht genes. It appears that the

difference in FHB resistance between these two varieties and the others on the UK

National List of 2003 may, in large part, be simply a reflection of the presence or

absence of Rht-D1b. Under conditions of moderate disease pressure, use of the Rht-

B1b semi-dwarfing allele may provide the desired crop height without compromising

resistance to FHB to the same extent as lines carrying Rht-D1b.

In the Frontana-derived population we identified a major stable QTL on

chromosome 2B and one on 3A that was only effective in field conditions with low

disease pressure. Similar QTL have been observed in Frontana, indicating that they

retain their efficacy in different genetic backgrounds.

Disease among 30 varieties was correlated following spray inoculation with F.

culmorum and M. majus although M. majus was much less aggressive. In contrast,

following point inoculation, M. majus was unable to spread beyond the infected

spikelet whereas the DON-producing F. culmorum isolate spread into the rachis and

throughout the head. Symptoms produced by M. majus, a non toxin-producing species

are almost identical to those produced by Tri5-transformants, being restricted to single

spikelets and unable to spread throughout the spike.

Experiments comparing NIV and DON producing isolates of F. graminearum

showed that, while the two isolates caused similar levels of disease initially, the DON

producer spread much more rapidly in the spike than the NIV producer. We concluded

that use of appropriate virulent NIV chemotype isolates of F. graminearum might be

used in spray inoculation trials to determine relative levels of Type I resistance in

wheat. To complement this, point inoculation with a virulent DON-producing isolate

13

can be used to evaluate levels of type II resistance. We used such isolates to identify

FHB resistance within a large collection of wheat lines from the International Wheat

and Maize Centre (CIMMYT), Mexico. Several lines exhibiting high levels of FHB

resistance in field trials were found to also possess high levels of Type II resistance

following point inoculation. More significantly, a few lines exhibited high levels of FHB

resistance in field trials, but were highly susceptible to point inoculation indicating that

their resistance is predominantly of Type I. Only a few sources of this type of

resistance have been identified to date due, in large part, to the greater technical

challenges associated with the unequivocal identification of Type I resistance. We

propose that the use of appropriate non DON-producing FHB species or isolates in

spray inoculation trials combined with point inoculation using DON-producing isolates

will greatly aid the identification and characterisation of wheat for Type I and Type II

resistance to FHB.

14

FULL PROJECT REPORT

1 Introduction

Fusarium ear blight (FEB), more commonly termed Fusarium head blight (FHB), is a

damaging disease of wheat and barley in many cereal-growing areas of the world. The

disease has become devastating in the mid-west of USA, resulting in losses of well

over $1 billion over the last decade. The disease can be caused by a number of fungi,

chiefly Fusarium species. Whereas F. graminearum is predominant worldwide, it was,

until recently, not often found in the UK where the main pathogens were F. culmorum

and Microdochium nivale. However, this species has now become established across

the UK and is now frequently isolated from wheat ears (Turner et al, 1999). As well as

causing loss of yield and quality this disease is of major concern because of the

production of mycotoxins by many of the Fusarium species responsible (M. nivale does

not produce mycotoxins). The chief mycotoxins of concern are trichothecenes and the

most prevalent in cereals is deoxynivalenol (DON) although a derivative nivalenol

(NIV) is also of relevance to the UK and Europe.

To date, the disease has occurred only sporadically in the UK. However,

observed changes in the species detected in the UK indicate an increasing risk. In

addition, altered climatic conditions in the future may lead to an increase in the

incidence of this disease, irrespective of changes in the pathogen species responsible.

In an effort to ensure continued and improved health of consumers the EU recently

introduced legislation setting limits to the amount DON mycotoxin permitted in grain

and wheat products. It is imperative that the UK is positioned to comply with any such

legislation and maintain consumer confidence through ensuring a disease and toxin

free crop. It is widely accepted that host resistance is an essential component of any

attempt to reduce the level of disease and associated mycotoxins.

The problem addressed in the REFAM project was that progress in breeding

varieties resistant to Fusarium species and at low risk of mycotoxin accumulation is

currently constrained because very little is known about the genetics of resistance.

Efforts to study resistance are hindered because of the quantitative nature of

the resistance with several genes being required to provide significant protection.

Disease symptoms do not always appear to be related to toxin accumulation. Although

this may partly reflect the inadequacy of the assessment techniques (Nicholson et al,

unpublished) it is clearly important to assess resistance to both disease and

mycotoxin accumulation .In the field, the disease consists of a complex of different

species (some producing toxins while others do not) and it remains unclear whether

15

resistance to toxin producing species also always confers resistance to species such as

M. nivale that do not produce toxins. It is highly likely that the effect will depend upon

the mechanism(s) behind a particular resistance and, hence, depends upon the wheat

variety involved.

Wheat heads differ in susceptibility as they mature, being most susceptible at

mid-anthesis. In artificially inoculated tests, failure to apply the fungus at the correct

time will lead to escape and prevent assessment of true resistance. In the field, other

characters, such as plant height and ear morphology may influence humidity and

hence infection. This will also interfere with attempts to distinguish true resistance

from disease escape in a particular environment. Relatively few researchers have

addressed these problems because of the requirement for significant investment in

personnel and infrastructure.

The genetic basis for resistance has been investigated for only a few varieties.

The best studied of these is the Chinese variety “Sumai 3” in which it is thought that a

major part of the resistance is due to a factor on the short arm of chromosome 3B.

Resistance from this variety has yet to be incorporated into backgrounds suitable for

growth in the UK. There is little published information on the genetic basis of other

potential FHB resistance sources although moderate levels of resistance have been

identified in UK wheat varieties (Nicholson et al., unpublished) The genetic basis for

the resistance of these varieties is not known.

It is important that further resistance sources are identified, and the genetic

basis of resistance understood, in order to complement that of Sumai 3 or its

derivatives. Reliance upon one resistance factor of a single variety would pose a risk

to the long term security of the wheat crop in the UK and elsewhere.

For barley, even less is known than for wheat about the status of FEB resistance, as

varieties are not routinely screened for resistance during breeding or official trials.

There is an urgent need to examine current UK and European material for resistance

to Fusarium spp. / M. nivale and mycotoxin production, in order to establish what

resistance is available for immediate exploitation, before proceeding to search more

widely for sources of improved resistance.

The purpose of the REFAM project was to identify and characterise the best

available sources of FHB resistance in wheat and barley and develop rigorous and

detailed knowledge of the genetics of FHB resistance in wheat. The approach will

facilitate long-term progress in the scientific study of resistance to the disease and

toxin accumulation and enable the development of efficient marker assisted selection

(MAS) procedures within breeding programmes of the companies that are partners in

16

this application.

Advances in molecular marker technology, combined with statistical methods to

analyse quantitative traits, such as FHB resistance, provide an opportunity to

significantly improve the understanding of the genetic basis of FHB resistance. In

recent work funded by the EU and DEFRA, the JIC has demonstrated that a rigorous

approach to disease assessment and analysis of the traits that form components of

the overall resistance can be used to investigate the genetic basis of FEB resistance.

This approach allows the dissection of resistance and identification of molecular

markers linked tightly to the resistance components.

Plant breeders require more information about the genetics of resistance to

FHB, particularly in respect to the relationship between disease and mycotoxin

accumulation as well as the relative efficacy of resistance towards toxin producing and

non-toxin producing FHB species. There is a scarcity of sources of FHB resistance and

this project will contribute towards the needs of breeders through the identification

and characterisation of new sources of resistance. This information will help them to

accumulate genes from different sources (and with differing chromosomal locations),

in order to build up effective, durable resistance in their breeding programmes.

The project consisted of three work packages (WP) with associated objectives.

1a Evaluation of UK RL and selected European winter wheat and barley lines with

known FHB resistance.

1b Comparison of resistance efficacy against toxin producing and non-toxin

producing FHB species by spray and point inoculation.

1c Identification and characterisation of new sources of FHB resistance by

screening wheat collections held at the JIC.

2 Analysis of resistance in populations from resistant/susceptible crosses. Three

populations were studied Spark x Rialto, Soissons x Orvantis and RL4137 x Timgalen.

3a Microsatellite (SSR), amplified fragment length polymorphism (AFLP) and

Diversity Array (DArT) markers were used to develop genetic maps of the selected

crosses. Quantitative trait locus (QTL) analysis was undertaken to determine the

precise location and relative magnitude of FHB resistance and the identification of

molecular markers tightly linked to FHB resistance.

17

3b Identification of molecular markers tightly linked to FHB/toxin resistance,

suitable for inclusion in marker assisted selection procedures within breeding

programmes.

2 Materials and methods

2.1 Plant materials

Recombinant inbred lines (RILs) of a Cross between RL4137 (TRL1) and Timgalen

(TRL2) and two doubled haploid (DH) populations: Spark/Rialto, Soissons/Orvantis,

was used in the study. In additions to the mapping populations, the near isogenic

lines (NILs) of Rht1 and Rht2 semi-dwarfing alleles in Mercia and Maris Huntsman

background (Flintham et al, 1997), 79 winter wheat varieties and 300 lines from

CIMMYT FHB resistance breeding programme were used in the study.

2.2 Genotyping

The genetic maps for the mapping populations were constructed using various

markers including simple sequence repeats (SSR), amplified fragment length

polymorphism (AFLP) (Vos et al, 1995) and Diversity Array Technology (DArTTM)

markers (Jaccoud et al, 2001). The DArT markers were added by The Triticarte Pvt.

Ltd., Australia. SSRs were visualised on Polyacrylamide silver staining gels or ran on

ABI sequencing machine and scored. The Rht1 and Rht2 perfect markers were

mapped as described in Ellis et al (2002). SSR and DArT markers permitted the

assignment of linkage groups (LGs) to chromosomes according to the previously

published wheat genetic maps

(http://www.triticarte.com.au/pdf/TriticartewhtmapalignV1-2.xls; Semagn et al,

2007; Somers et al, 2004).

2.3 Phenotyping for FHB resistance

The FHB phenotyping was carried out both in field and polytunnel experiments using

either using randomized complete blocks design or randomised incomplete block

design after classification of the genotypes according to flowering date. The field plot

size was double rows of 1 m length with 17 cm row spacing. The field trials were

conducted either at the John Innes Centre (JIC), Norwich, Nickerson Seeds Ltd,

Suffolk, Advanta Seeds (UK) Ltd., Norfolk, National Institute for Agricultural Botany

(NIAB), Cambridge, Central Science Laboratory (CSL), York or Elsoms Seeds Ltd.,

Lincolnshire. All the polytunnel experiments were conducted at JIC. Polytunnel

experiments used 12 plants of each line grown separately in 1 litre pots of John Innes

18

potting mix, with one seed per pot and grown in polytunnel. The pots containing

individual plants were arranged in a randomised complete block design, prior to

inoculation. Plants were spray inoculated at mid-anthesis (growth stage 65 (Zadoks et

al, 1974) with a conidial suspension (1 x 105 spores ml-1) of highly virulent DON

producing isolate Fusarium culmorum (Fu42 or UK1) or a NIV producing isolate of F.

grameniarum (F86). The inoculum preparation, plant husbandry, trial set-up and

disease assessment were carried out as described in Gosman et al (2005). All the

inocula were amended with 0.05 % Tween 20 and the inoculations were carried out in

the evening.

The variety Mercia was assessed for resistance to a DON-producing isolate UK1

and a NIV-producing isolate F86 of F. graminearum following spray and point

inoculation. Two experiments were performed in an unheated polytunnel with different

inoculum loads. In the high titre trial, conidial suspensions were 1 x 105 ml-1 and 1 x

106 ml-1 for spray and point inoculations, respectively. For the low titre trial, conidial

suspensions were 1 x 104 ml-1 and 1 x 105 ml-1 for spray and point inoculations,

respectively.

Disease was assessed as the percentage (0 to 100 %) of visually infected

spikelets. In the polytunnel trials, disease was assessed several times at intervals of 7

days from the date of post inoculation and area under the disease progress curve

(AUDPC) was calculated (Buerstmayr et al, 2000) and used in analysis. In addition to

disease, the morphological traits such as awns, plant height (PH) and weight of

infected spikelets (WIS) were recorded.

2.4 DON analysis, DNA extraction and quantitative PCR

DON was extracted from milled grain samples using 10% methanol (10 ml g-1) and

stored at -20 0C prior to analysis. An appropriate dilution in distilled water was made

from each sample and 50 μl aliquots assessed for DON content using the Ridascreen®

Fast DON™ (R-Biopharm Rhône Ltd.) enzyme linked immuno-assay (ELISA) according

to the manufacturer’s instructions. Absorbance was measured at 450 nm using a

Wallec 400 plate reader and the sample data converted to DON concentration by

reference to a standard curve generated from DON standards provided in the kit.

DNA extraction and competitive PCR were performed as described previously (Gosman

et al, 2004). Quantitative PCR analysis using F. culmorum specific primers was carried

out according to the method of Nicholson et al (1998) and the amount of fungal DNA

was expressed as a percentage of the total DNA content of the sample as described by

Gosman et al (2004).

19

2.5 SSR haplotyping

Comparison of microsatellite allele size with FHB resistant and susceptible cultivars

from Asia, Europe and USA described in published QTL mapping studies were used to

infer the origin of FHB resistance QTL in the trial varieties and to identify potentially

novel loci. The microsatellite loci analysed were reported in those studies to flank FHB

resistance QTL on chromosomes 1B, 3B, 5A and 6B. SSR loci Xgwm389, Xgwm533

and Xgwm493 (3BS) and Xgwm293, Xgwm156 and Xgwm304 (5A) were used to

identify QTL Qfhs.ndsu-3BS and Qfhs.ifa-5A derived from Sumai-3 (Anderson et al,

2001; Buerstmayr et al, 2003) (Table 2.1). The presence of a QTL in a similar location

to Qfhs.ifa-5A identified in the Romainian cultivar Fundulea F201R (Shen et al, 2003)

was monitored using Xgwm304. Two QTL on 5AL of Renan (Gervais et al, 2003) were

identified using Xgwm639, Wmc415 and Barc151 for QFhs.inra-5a2 and Wmc524 and

Xgwm410 for QFhs.inra-5a3. The presence of the ω-secalin gene and microsatellite

markers Xbarc008 and Xgwm018 were used to identify the QTL on 1B derived from

Fundulea F201R (Shen et al, 2003). As an additional reference, the Russian cultivar

Aurora was included in the analysis as it is likely that F201R inherited the 1RS-1BL

translocation from Aura, which has Aurora (carrier of the 1RS-1BL translocation) in its

parentage (Shen et al. 2003). On 6BS an interval was identified which encompassed

QTL in Sumai-3 (Wmc105) and Ning 894037 (Xgwm088) reported by McCartney et al.

(2004) and Shen et al (2003b), respectively. Additional loci (Xgwm133 and Barc198)

identified from the consensus wheat map of Somers et al (2004) were used as

flanking markers for this QTL. Polymorphic information content, or PIC values were

calculated using the formula of Botstein et al (1980).

The sources of primer sequences for microsatellites are identified by the marker

prefix, e.g. (Xgwm) published in Röder et al (1998), (Barc) published in Song et al

(2002) and (Wmc) published by Somers & Isaac (2004). PCR amplification for SSR

analysis was carried out in 6.25μl volumes using Qiagen HotStarTaq Mastermix

(Qiagen), primers at 0.2μM; one of which was fluorescently labelled with FAM, VIC,

NED, PET or HEX (Applied Biosystems) using approximately 20ng of genomic DNA.

PCR was performed using a MJ Research Tetrad Thermocycler (MJ Research). The

reaction mixtures were denatured at 95 0C for 15 minutes, followed by 34 cycles

consisting of 94 0C for 1 minute, primer dependent annealing temperature for 1

minute (ramp speed 0.5 0C sec-1), and extension at 72 0C for 1 minute. Amplification

was completed with a 10 minute extension at 72 0C and held at 10 0C.

20

Table 2.1 Summary of chromosomal locations of FHB resistance QTL on chromosomes 1B, 3BS, 5A and 6B, flanking markers and references to published studies

Chromosome Cultivar Markers(s) Reference 1B Fundulea F201R Barc008, Xgwm018 Shen et al.(2003a) 1B CM82036 Glu-B1 Buerstmayr et al (2002) 3BS Sumai-3 Xgwm389, Xgwm533, Xgwm493 Anderson et al (2001) 3BS Ning 7840 Xgwm389, Xgwm533 Guo et al. (2003) 3BS CM82036 Xgwm533, Barc133, Xgwm493 Buerstmayr et al (2003) 3BS Ning 894037 Xgwm389, Xgwm533, Xgwm493 Shen et al (2003b) 3BS Huapei 57-2 Xgwm389,Barc133, Xgwm493 Bourdoncle & Ohm (2003) 5A CM82036 Xgwm293, Xgwm304, Xgwm156 Buerstmayr et al (2003) 5A Fundulea F201R Xgwm304 Shen et al (2003a) 5A Sumai-3 B1 awning suppressor Ban & Suenaga (2000) 5A-1 Renan Xpsr0170a Gervais et al (2003) 5A-2 Renan Xgwm639b Gervais et al (2003) 5A-3 Renan B1 awning suppressor Gervais et al (2003) 6B Ning 894037 Xgwm088, Xgwm644 Shen et al (2003b)

Hierarchical cluster analysis was performed using the Euclidian test with single

link as the clustering method. The similarity matrix was calculated using allele size

data generated from SSR loci associated with the six FHB resistance QTL described

previously.

2.6 Statistical analysis

All the statistical analysis was performed using GenStat for Windows 9th edition (copy

right Lowes Agricultural Trust, Rothamsted Experimental Station, UK). Analysis of

variance (ANOVA) was carried out using the generalised linear model (GLM) of

regression analysis. Broad sense heritability (h2) was estimated from the ANOVA using

the formula: h2 = σG2/[σG

2 + (σe2/r)], with σG

2, the genetic variance; r, the number of

replicates per genotype (Nyquist, 1991).

In the current project, linkage maps were constructed for RL4137 (TRL1) X

Timgalen (TRL2) and Soissons X Orvantis mapping populations using JoinMap (version

3.0) (Van Ooijen & Voorrips, 2001) and the map distances were calculated using

Kosambi mapping function (Kosambi 1944). The QTL detection was carried out using

MapQTL® 4.0 (Van Ooijen, 2004). First, for each trait, Krusskall-Wallis test was

performed to detect the association between markers and traits individually. In a

second step, interval mapping (IM) was performed to identify the major QTL.

Automatic cofactor selection was used to fit the multiple QTL model (MQM) (backward

elimination (P>0.02)) and detect significantly associated markers as cofactors. For

each trait, a 1000X permutation test was performed to identify the LOD threshold

corresponding to a genome-wide false discovery rate of 5 % (P<0.05). The QTL

21

detected above the LOD threshold that explained more than 10 % of the variance (R2)

in at least one environment/experiment were classified as major QTL and those

explaining less than 10 % as minor QTL. Linkage maps were drawn using MapChart

(Vorrips, 2002).

22

3 Evaluation of UK released varieties and selected European winter

wheat lines with known FHB resistance.

Fusarium head blight (FHB) of wheat is caused by several fungal species that cause

similar symptoms. Fusarium graminearum is the major pathogen worldwide, while F.

culmorum tends to predominate in maritime regions (McMullen et al, 1997; Windels,

2000). F. avenaceum and F. poae are also frequently associated with FHB, particularly

in Northern Europe. In addition to the true Fusarium species, two Microdochium

species M. majus and M. nivale also cause FHB and are particularly prevalent where

cooler, wetter conditions prevail such as the UK (Lemanczyk & Sadowski, 2002).

FHB is of particular concern because many of the Fusarium species produce

mycotoxins that contaminate infected grain and pose a risk to human and animal

consumers (Pestka & Smolinski, 2005). The most common mycotoxin in blighted grain

is the trichothecene deoxynivalenol (DON), produced by F. graminearum and F.

culmorum. A second, closely related trichothecene, produced by certain isolates of

these species is nivalenol (NIV).

The use FHB-resistant cultivars is the most cost-effective and environmentally

pragmatic approach to manage this devastating disease. FHB resistance is

quantitatively inherited and is heavily influenced environment during anthsis. To date

no variety has been found to be immune to FHB (Snijiders, 1990). Advances in

phenotyping, combined with statistical methods to detect quantitative trait loci (QTL)

have led to the identification of numerous QTL for FHB resistance in spring wheat

varieties which could be introgressed into elite agronomically superior cultivars (Van

Sanford et al, 2001). Several sources of FHB resistance genes have been identified

among Chinese, and Chinese derived spring wheat notably Sumai 3 has already been

widely deployed in breeding programs worldwide.

Despite the existence of a significant amount of FHB resistant European winter

wheats, resistance in only a few varieties has been genetically characterised to date

(Paillard et al, 2004; Scmolke et al, 2005; Shen et al, 2003b). In contrast to the

resistance in some spring wheat varieties, FHB resistance in winter wheat germplasm

appears to be due to numerous QTL of moderate to small effect. Microsatellite

markers can be used to infer the origin and genetic relationship between FHB

resistance QTL which could be used to identify appropriate parental combinations in

order to ‘pyramid’ previously characterised resistance genes and to identify potentially

novel sources of resistance (McCartney et al, 2004). Diversification of resistance

sources in breeding programs should reduce the risk of the emergence of virulent

pathogen strains.

23

Due to the previously low incidence of FHB in the UK, a comprehensive

assessment of resistance in elite UK winter wheat germplasm has not yet been

undertaken. One aim of the current study was to assess the FHB reaction of winter

wheat varieties currently on the UK National List (NIAB, Cambridge) and to identify

the presence of characterised FHB QTL and infer potentially novel sources of

resistance through allele size comparison of microsatellite loci associated with QTL in

Chinese and Chinese derived germplasm.

3.1 Results

3.1.1 Fusarium head blight resistance

In (2003) 72 winter wheat varieties from the UK National List (NIAB), plus additional

entries from France and Germany were phenotyped for resistance to FHB (Table 3.1).

Trials were carried out at two field sites in the UK; one in the South East of England at

NIAB in Cambridge, and the other in the North of England at CSL in York. A third trial

was carried out in a high temperature and humidity facility (polytunnel) at the JIC in

Norwich. Relative to the most susceptible variety at each site, average disease levels

were highest at JIC (56.35%) followed by NIAB (28.65%) and CSL (15.34%).

Coefficients of correlation between sites were low for the relationship between JIC and

NIAB (0.36, P = 0.002) to high for the relationship between JIC and CSL (0.40, P =

0.001) and between NIAB and CSL (0.52, P < 0.001). Analysis of variance (ANOVA)

indicated that genotypic variance was highly significant (P < 0.001) and within year

heritability calculated from variance component estimates over all sites was medium

sized (H2 = 0.57). Comparison between the average score of equal sized groups of

varieties from the UK and control entries from Continental Europe over all three sites

indicated that UK varieties were significantly (t = -3.05, P = 0.005) more susceptible.

Across sites, relative to the cultivar Wizard (which in previous tests had been

shown to be FHB susceptible), the UK variety Macro was the most susceptible (highest

post inoculation score of 35% infection) and the most resistant was the French

cultivar Nirvana (highest post inoculation score of 2.4 % infection). The most resistant

varieties, all with significantly (P ≤ 0.05) less visual disease than Wizard, were (most

resistant first) Nirvana, Renan, Tambor, Petrus, Soissons, Vector, Atlantis, Centrum,

Ornicar, Spark and Piko (Figure 3.1). All these varieties, with the exception of Vector

and Spark, were from Continental Europe. The most susceptible varieties were (most

susceptible first) Macro, Darwin, Tanker, Scorpion 25, Goodwood, Xi19, Warlock,

24

Riband, Einstein and Bentley (Figure 3.1). With the exception of Bentley, all these

varieties were from the UK.

Table 3.1 Origin and pedigree information of 79 winter wheat trial entries comprised of 7 FHB resistant reference varieties used in previous QTL studies (REF), 41 varieties undergoing recommended list trials, or were on the UK recommended list in (2003) (RL)1, 12 UK national list varieties (NL) and 19 control varieties which have previously been tested for FHB resistance by collaborative partners (CON) Variety / line code Breeder / seed source Status Country of

origin Sumai-3 Agriculture and Agri-food Canada REF China Wuhan 2-37E John Innes Centre REF China CM82036 CIMMYT REF Mexico Aurora CIMMYT REF Russia Fundulea F201R Institute for Cereals & Industrial Crops

- Fundulea REF Romania

WEK0609 Pioneer HiBred Seeds REF USA Renan INRA REF France 01ST2031 Monsanto CON France A42-02 Advanta NL UK A48-02 Advanta NL UK Access CPB Twyford RL UK Apache Monsanto CON France Arran Nickerson RL UK Atlantis Nickerson CON Germany Bandit Nickerson NL UK Batis Monsanto CON Germany Bentley Desprez RL France Biscay CPB Twyford RL UK Carlton Elsoms NL UK Centrum Monsanto CON France Charger Monsanto NL UK Claire Nickerson RL UK Consort Monsanto RL UK Contra Monsanto CON Germany Cordiale (CPBT W83) CPB Twyford RL UK CPBT W87 CPB Twyford RL UK CWW 00/22 Monsanto RL UK Dart (CWW 00/47) Monsanto RL UK Darwin Nickerson NL UK Deben Nickerson RL UK Dekan Nickerson CON Germany Dick Cebeco RL UK Einstein Nickerson RL UK ELS 00/21 Elsoms RL UK Excellence Monsanto NL UK German A Monsanto CON Germany Gladiator (CWW 00/33) Monsanto RL UK Goodwood Cebeco RL UK Grief Monsanto CON Germany Hereward Monsanto RL UK Istabraq (NSL WW47) Nickerson RL UK Macro Monsanto NL UK Malacca CPB Twyford RL UK

1 Information provided by the (2003) NIAB Recommended List and UK National list

25

Variety / line code Breeder / seed source Status Country of origin

Napier Monsanto RL UK Nirvana Monsanto CON France Nijinsky (NSL WW46) Nickerson RL UK Option Monsanto RL UK Ornicar Monsanto CON France Orvantis Monsanto CON France Pennant Elsoms NL UK Petrus Monsanto CON Germany Piko Monsanto CON Germany Rialto Monsanto NL UK Riband Monsanto RL UK Richmond Cebeco RL UK Robigus CPB Twyford RL UK Romanus Nickerson CON Germany Savannah Advanta RL UK Scorpion 25 Advanta RL UK Senator (CPBT W90) CPB Twyford RL UK Skater Nickerson RL Germany Smuggler (A30-00) Advanta RL UK Soissons Desprez RL France Solstice Advanta NL UK Spark Nickerson NL UK Steadfast (CWW 00/40) Monsanto RL UK SW Tataros Hadmersleben RL Germany Tambor Monsanto CON Germany Tanker Elsoms RL UK Tellus CPB Twyford RL UK Travix Nickerson CON UK Vector Advanta RL UK Vergas Nickerson CON Germany Warlock 24 Advanta RL UK Welford Elsoms RL UK Winnetou Nickerson CON Germany Wizard CPB Twyford RL UK Xi19 Advanta RL UK Zebedee (A45-02) Advanta NL UK 1 Information provided by the (2003) NIAB Recommended List and UK National list

26

Figure 3.1 (A) Visual disease over three sites, (B) visual disease at the NIAB trials site (C) deoxynivalenol (DON) (ng / mg) and (D) fungal DNA (FDNA) (LOG10 fungal DNA as a % of total) content of grain from 72 wheat lines. Values are average deviation from the highly FHB susceptible variety Wizard. All entries are ordered (greatest relative disease reduction on the left) according to the mean deviation from Wizard at the NIAB site (which provided the samples for DON and FDNA analysis)

27

3.1.2 Relationship between FHB symptoms and the deoxynivalenol

(DON) and fungal DNA content of grain

In addition to visual disease, grain samples from the NIAB site were assessed for the

trichothecene deoxynivalenol (DON) and for fungal DNA (FDNA) content. The average

DON content of samples was 1.07 mg kg-1 with a range of < 0.01 mg kg-1 to 3.75 mg

kg-1 and the average FDNA level was 0.031%, with a range of 1.72 x 10-4% to 0.38%.

The entries with the lowest toxin content were French varieties Soissons and

01ST2031 and the entry with the highest was the French variety Orvantis. The entry

with the lowest FDNA content was the UK variety Spark and the one with the highest

was the UK variety Richmond (Figure 3.1). Analysis of traits relative to the FHB

susceptible cultivar Wizard indicated that estimates of disease severity sometimes did

not correspond with DON and FDNA levels in grain samples. For example, 01ST2031

and Soissons showed only moderate resistance at NIAB but both had the lowest toxin

levels in the trial. Similarly, grain samples from the moderately resistant French

cultivar Renan had a level of FDNA which was non-significantly (P > 0.05) different to

entries with the highest levels of visual disease (Figure 3.1). However, coefficients of

correlation were high (0.62, P < 0.001) for the relationship between visual disease at

NIAB (FHB-NIAB) and DON and moderate between FHB-NIAB and FDNA (0.36, P =

0.01) (Table 3.2).

Table 3.2 Coefficients of correlation for the relationship between visual disease (% damage) over sites (FHB-OS), visual disease (percent damage) at NIAB (FHB-NIAB), deoxynivalenol (DON) and fungal DNA (FDNA) content of grain samples from the NIAB site. FHB-OS FHB-NIAB FDNA FHB-NIAB 0.61 *** -- -- FDNA (%) 0.53 *** 0.36 ** -- DON (mg kg-1) 0.45 *** 0.62 *** 0.61 *** *** Significantly different from zero at P < 0.001 ** Significantly different from zero at P < 0.01

3.1.3 Haplotyping with Known SSRs linked to FHB resistance QTL

A total of 17 microsatellite markers associated with six FHB QTL were used to

genotype 83 wheat varieties (72 entries plus 11 reference cultivars) (Table 4). The

polymorphic information content (PIC) values of the SSR loci ranged from 0.38 to

0.86 (mean = 0.65). Markers detected from four to 12 (mean = 7.35) alleles, giving

rise to 83 unique haplotypes. Hierarchical cluster analysis was used to group lines

based on marker haplotype; reference cultivars are shown in bold type (Figure 3.2).

28

Varieties grouped together into five main clusters the composition of which largely

reflected their country of origin. The largest cluster was group 2 which consisted of

mainly UK germplasm (21 entries) with a few (4 entries) varieties from France and

Germany. Group 1 (6 entries) consisted of FHB resistant germplasm from France and

Germany. Groups 3 and 4 (8 and 7 entries, respectively) consisted of only UK

varieties, and group 5 consisted of the FHB resistant reference varieties, Wuhan 2-

37E, CM82036, WEK0609 and Fundulea F201R. The Russian variety Aurora was not

closely related to any of the other entries, and the most divergent entry was the

German variety Petrus.

Interval length and haplotype diversity (total number of haplotypes and number

of haplotypes per centi-Morgan (cM)) was assessed for each QTL interval. Relative to

the other chromosome regions, 6BS and 1B had the shortest interval length (Table 4).

In terms of haplotype number, QFhs.inra-5a2 had the highest haplotype diversity;

however, 6BS had the highest number of haplotypes per cM. By contrast, Qfhs.ifa-5A

had the longest interval length and one of the lowest levels of haplotype diversity.

The allele distribution of the 17 microsatellite loci, grouped according to QTL

interval, together with the FHB reaction for each variety is presented in Table 5. The

most resistant (Nirvana) and susceptible (Macro) varieties in trial were used as

standards in regression analysis at the 0.05% level of significance to classify entries

into resistant, susceptible and moderate FHB resistance categories.

At the 1B locus the FHB resistant entry Petrus (German) and FHB resistant

reference varieties Wuhan 2-37E, CM82036, Fundulea F201R and WEK0609 all carried

the 1RS-1BL translocation and shared exactly the same haplotype as Aurora.

However, varieties with a range of reactions to FHB carried the 1RS-1BL rye

translocation and had the same allele size as Aurora at the Barc008 locus; these were

A42-02, Access, Atlantis, ELS00/21, Excellence, Gladiator, Napier, Pennant, Rialto,

Savannah, Senator, Steadfast, Travix, Vector, Vergas and Welford. There were no

alternative haplotypes consistently associated with resistant or moderately resistant

entries. The most common haplotype was (numeric values are base pairs) Xgwm018

(196), 1RS-1BL absent, Barc008 (257) with a total of 15 occurrences. This haplotype

was shared by a group of varieties of diverse origin with a range of FHB reactions

(Table 3.3).

29

Table 3.3 Allele sizes of microsatellite loci associated with FHB resistance QTL from wheat varieties on the United Kingdom (2003) National List plus reference cultivars from China, Mexico, USA and continental Europe. Within each interval, loci have been ordered (markers closest to, or on the short arm on the left and markers closest to, or on the long arm on the right) according to their map position in the consensus wheat map published by Somers et al (2003). Numeric values are amplicon sizes in base pairs for each marker, nd is no data. SSR alleles associated with FHB resistance QTL identified in previously published studies are colour coded: yellow identifies entries with the same size alleles as the Sumai-3 6BS resistance QTL, red identifies entries with the same size alleles as CM82036 linked to the Qfhs.ndsu-3BS (3BS) and Qfhs.ifa-5A (5A) QTL, blue identifies entries with the same size alleles as the Aurora 5A QTL, purple identifies entries with the same size alleles as Renan linked to the QFhs.inra-5a2 (5AL (2)) and QFhs.inra-5a3 (5AL (3)) QTL. The presence of rye chromatin associated with the 1RS-1BL translocation was identified using the duplex PCR assay of De Froidmont (1998), Pres 1RS-1BL present, Abs 1RS-1BL absent.

30

At Qfhs.ndsu-3BS there were no entry genotypes that matched any of the CM82036

alleles; however, reference varieties Sumai-3, Wuhan 2-37E and WEK0609 shared the

same haplotype as CM82036. There were no alternative haplotypes consistently

associated with resistant or moderately resistant entries. The most common haplotype

was Xgwm389 (138), Xgwm533 (Null), Xgwm493 (146) with 13 occurrences. This

haplotype was shared by moderately resistant and susceptible varieties of UK origin.

As previously mentioned, two QTL associated with CM82036 (marked red) and

F201R (marked blue) have been identified at the Qfhs.ifa-5A locus. In the current

study, there were no entry genotypes with a haplotype that exactly matched either

the CM82036 or F201R haplotype. Among the reference cultivars, Sumai-3 and Wuhan

2-37E matched the CM82036 haplotype and WEK0609 and Renan matched that of

Aurora but not that of F201R which differed from the other three varieties at the

Xgwm156 locus. An additional 5A related haplotype of Xgwm293 (174), Xgwm304

(195) Xgwm156 (325) was, with the exception of Option, associated with resistant

and moderately resistant entries from the UK and continental Europe; these were,

Gladiator, Napier, Access, Biscay, Nirvana, Batis, Pennant, Option and Tellus. FHB

susceptible UK varieties, Dick and Darwin matched with Aurora at the Xgwm293 and

Xgwm156 loci, and the FHB resistant reference variety F201R (which has Aurora in its

genealogy) matched with Aurora at the Xgwm293 and Xgwm304 loci. The most

common haplotype was Xgwm293 (174), Xgwm304 (195), Xgwm156 (310) with 29

occurrences. This haplotype was shared by a group of varieties of diverse origin with a

range of FHB reactions.

At QFhs.inra-5a2 there were no entry genotypes or reference cultivars that

matched the Renan haplotype. There were no alternative haplotypes consistently

associated with resistant or moderately resistant entries. The most common haplotype

was Xgwm639 (134), Wmc415 (140), Barc151 (246) with four occurrences. This

haplotype was found among susceptible UK varieties.

At QFhs.inra-5a3 there were no entry genotypes or reference cultivars that

exactly matched the Renan haplotype. However, resistant and moderately resistant

entry varieties Apache, Carlton, ELS 00/21, Pennant and Welford had the same allele

as Renan at the Xgwm410 locus and resistant entry varieties Soissons and 01ST2031

and reference cultivars Sumai-3, WEK0609, CM82036, Fundulea F201R, and Wuhan

2-37E had the same allele as Renan at the Wmc524 locus. There were no alternative

haplotypes consistently associated with resistant or moderately resistant entries. The

most common haplotype was Wmc524 (172) and Xgwm410 (347) with 32

31

occurrences. This haplotype was shared by a group of varieties of diverse origin with a

range of FHB reactions.

At the 6BS interval, with the exception of Wuhan 2-37E which matched exactly,

none of the other varieties had the same allele size as Sumai-3 at any of the loci. An

additional haplotype of 134, 350, 127 at the Xgwm088, Wmc105 and Barc198 loci

respectively, was associated with the resistant and moderately resistant entry and

reference varieties which were all, with the exception of Pennant, from continental

Europe; these were ELS 00/21, Aurora, Renan, F201R, SW Tataros, Tambor, Pennant.

The most common haplotype was Xgwm133 (190), Xgwm088 (124), Wmc105 (336),

Barc198 (162) with 18 occurrences. This was shared by a group of susceptible and

moderately resist.

32

Coefficient of similarity Figure 3.2 Dendrogram produced by hierarchical cluster analysis of 72 trial lines and 7 reference varieties using the Euclidian test with single link as the clustering method. The similarity matrix was calculated using allele size data from 17 microsatellite loci associated with six FHB resistance QTL. Shaded areas identify major groupings.

Vergas

Vector

Travix

Tellus

Tanker

Tambor

SW Tataros

Spark

Solstice

SoissonsSkater

Scorpion 25

Savannah

Romanus

Robigus

Richmond

Riband

Rialto

Piko

Petrus

Pennant

Orvantis

Ornicar

Option

Istabraq

Nijinsky

Nirvana

Napier

Malacca

Macro

Hereward

Grief

Goodwood

GermanA

Excellence

ELS 00/21

Einstein

Dick

Dekan

Deben

Darwin

Dart

Wizard

Steadfast

Welford

Gladiator

CWW 00/22

Senator

CPBT W87

Cordiale

Contra

Consort

Claire

Charger

Centrum

Carlton

Biscay

Bentley

Batis

Bandit

Atlantis

Arran

Apache

Access

A48-02Zebedee

XI19

A42-02

Warlock 24

Smuggler

01ST2031

Renan

WEK0609F201-R

Aurora

CM82036Wuhan 2-37E

Sumai-3

Winnetou

0.880.92 0.901.00 0.98 0.96 0.94

1

2

3

4

5

Vergas

Vector

Travix

Tellus

Tanker

Tambor

SW Tataros

Spark

Solstice

SoissonsSkater

Scorpion 25

Savannah

Romanus

Robigus

Richmond

Riband

Rialto

Piko

Petrus

Pennant

Orvantis

Ornicar

Option

Istabraq

Nijinsky

Nirvana

Napier

Malacca

Macro

Hereward

Grief

Goodwood

GermanA

Excellence

ELS 00/21

Einstein

Dick

Dekan

Deben

Darwin

Dart

Wizard

Steadfast

Welford

Gladiator

CWW 00/22

Senator

CPBT W87

Cordiale

Contra

Consort

Claire

Charger

Centrum

Carlton

Biscay

Bentley

Batis

Bandit

Atlantis

Arran

Apache

Access

A48-02Zebedee

XI19

A42-02

Warlock 24

Smuggler

01ST2031

Renan

WEK0609F201-R

Aurora

CM82036Wuhan 2-37E

Sumai-3

Winnetou

0.880.92 0.901.00 0.98 0.96 0.94 0.880.92 0.901.00 0.98 0.96 0.94

1

2

3

4

5

33

3.2 Discussion

3.2.1 FHB status of UK winter wheat varieties

A total of 72 winter wheat varieties were phenotyped in the UK for FHB resistance in

the summer of 2003. Trial entries included 53 varieties on the UK National List (NIAB,

Cambridge) plus 19 check cultivars from France and Germany in which the FHB

reaction had previously been characterised by project collaborators. FHB resistance

was assessed by spray inoculation with a single aggressive isolate of F. culmorum

which is appropriate since there is no evidence that resistance to FHB is race or

species specific (Van Eeuwijk et al, 1995). Three contrasting environments, two field

sites at the Central Science Laboratory (CSL) and NIAB (low to moderate disease

levels respectively) and one controlled environment at the John Innes Centre (JIC)

(high disease level), were used to assess the stability of FHB reaction. In addition to

visual disease, the trichothecene mycotoxin, deoxynivalenol (DON) and fungal DNA

(FDNA) content of grain samples from the NIAB field site were assessed in order to

improve phenotype reliability and further characterise resistance. Although genotypic

variance was highly significant, environment and genotype by environment

components were also large. Similar sized environmental effects are often reported in

studies of FHB resistance in wheat mapping populations (Gervais et al, 2003;

Buerstmayr et al, 2003) and inbred lines (Buestmayr et al, 2004). Much of this

variance is probably due to variation in climatic conditions over years and between

sites (Mesterhazy et al, 1999). However, an unpublished study by Gosman et al

(2004) suggests that there can also be significant variation between winter wheat

cultivars for the period of optimal susceptibility to FHB. Assessment of visual disease,

DON and FDNA data from three inoculation timings indicated that some varieties had

a very narrow (three days or less) period of optimal susceptibility whereas others

remained highly susceptible over a six day period. These effects may, to some extent,

explain significant differences in disease response between sites observed in some

varieties.

Very few varieties on the UK National list had resistance levels which were

comparable those of resistant check cultivars. Of the 53 National List varieties tested,

only Soissons, Spark and Vector were significantly (P < 0.05) more resistant than the

FHB susceptible variety Wizard. In addition, even under the moderate disease

pressure at NIAB, a total of 24 National List varieties had DON levels which were

above the proposed EU action limits of 1.25 parts per million (ppm). These results

indicate that a significant effort will be required by the UK plant breeding community

34

to improve overall levels of FHB resistance. In general, increased resistance to

symptom development was correlated with reduced DON content of grain. Indeed,

certain resistant varieties (Tambor, Petrus and Soissons) and the moderately resistant

01ST2031 had particularly low toxin levels suggesting that they may have the ability

to degrade DON (Miller & Arnison, 1986) or reduce DON production by the pathogen.

However, some cultivars with significant levels of DON and FDNA (Dekan, Biscay,

Carlton, SW Tataros, Istabraq and Grief) appeared to be resistant suggesting that

visual disease may not be closely correlated to DON levels in some genotypes. A

comparative analysis by Paul et al (2005) suggests that the method of visual

assessment used may in part be responsible for lack of correspondence between DON

levels and symptom development in some instances. Over studies, these authors

found that DON content correlates better with assessments of damaged spikelets

within a head than with assessments of percentage infection within a plot.

Comparison of microsatellite (SSR) allele size was used to infer the origin of

FHB resistance and identify germplasm with potentially novel loci. From published

mapping studies, 17 SSR loci associated with six FHB resistance QTL on chromosomes

1B, 3B, 5A and 6B were selected. In accordance with haplotype analysis of wheat FHB

QTL by McCartney et al (2004) hierarchical cluster analysis of SSR allele size data

tended to group cultivars according to their country of origin producing three groups

of mainly UK germplasm and a small group of control varieties from France and

Germany. The most divergent variety was Petrus which had highly stable FHB

resistance over environments and significantly reduced both the toxin (DON) content

and fungal colonisation (FDNA) of grain. Petrus, therefore, appears to be a potent

source of novel FHB resistance which could be of great value in European breeding

programs.

In an attempt to identify potentially novel FHB resistance, comparison was

made with three loci of major effect on chromosomes 3BS, 5A and 6BS which have

been reported to condition resistance in the intensively studied cultivar Sumai-3. The

interval Qfhs.ndsu-3BS (3BS) associated with SSR loci Xgwm389, Xgwm533 and

Xgwm493 (Anderson et al 2001) has been reported in several studies where Chinese

or Sumai-3 derived germplasm has been the source of resistance (Buertmayr et al,

2003; Guo et al, 2003; Shen et al, 2003b). However, there was no allele size

homology between Sumai-3 and any of the test varieties in the current study

indicating that FHB resistance at the 3BS locus is absent among the germplasm

studied. Unsurprisingly, the Sumai-3 haplotype at Qfhs.ndsu-3BS was shared by the

Sumai-3 derivative CM82036. The Sumai-3 haplotype was also shared by the North

35

American wheat variety WEK0609 indicating that potent resistance identified in this

interval (Gosman et al, unpublished) may be allelic. Surprisingly, the Chinese variety

Wuhan 2-37E was identical to the Sumai-3 haplotype at all three QTL. This is in

contrast to McCartney et al (2004) who reported that Wuhan 2-37E had alternative

alleles at these loci. This result highlights the utility of selective genotyping in

identifying divergences in stocks that can arise in germplasm collections.

On 6BS, there was a similar absence of homology between test varieties for

SSR loci associated with major QTL for resistance in Ning 894037 (Shen et al, 2003b)

and Sumai-3 (McCartney et al, 2004). In the current study there was an alternative

haplotype which was, with the exception of the variety Pennant, associated with

resistant and moderately resistant germplasm of Continental European origin.

However, no QTL at this position was reported in studies of resistance in Renan

(Gervais et al, 2003) or F201R (Shen et al, 2003a) which also matched this haplotype

suggesting that this association maybe due to chance or that the putative QTL at this

locus was not segregating in the mapping populations used in the published studies

(Gervais et al, 2003; Shen et al, 2003a).

Two distinct haplotypes were identified at the Qfhs.ifa-5A (5A) locus, one

identical to CM82036 reported by Buestmayr et al (2003) shared by Sumai-3 and

Wuhan 2-37E, and the other identical to the Russian variety Aurora shared by

WEK0609 and Renan. However, F201R, which carries a potent QTL for resistance in a

similar position to Qfhs.ifa-5A (Shen et al, 2003a) differed at the Xgwm156 locus

suggesting that WEK0609 and Renan inherited this QTL from Aurora not F201R. No

other varieties matched either of the haplotypes suggesting that this locus was also

absent in the study germplasm. Interestingly, although Renan had a haplotype which

matched that of Aurora suggesting that it also has a QTL at this locus, no association

with FHB resistance has been reported (Gervais et al, 2003). However, potent FHB

resistance was associated with the Aurora haplotype in WEK0609 (Gosman et al,

unpublished) suggesting that the Aurora variant may not have been segregating in

the population used by Gervais et al (2003). In view of the relatively large genetic

distance delineated by Qfhs.ifa-5A it is also possible that the QTL may have been lost

in Renan through recombination without affecting the haplotype of the markers used.

According to the consensus map of Somers et al (2004), the order of markers at the

Qfhs.ifa-5A locus from the short arm is, Xgwm293, Xgwm304 and Xgwm156. The

haplotype of F201R differs from that of Aurora at the Xgwm156 locus, and the FHB

susceptible cultivars Dick and Darwin match Aurora at Xgwm156 and Xgwm293. This

result suggests that Xgwm293 and Xgwm156 may not be closely associated with

36

resistance in the Aurora variant of this locus indicating that loci more closely

associated with Xgwm304 might be needed for marker assisted selection (MAS). In

order that further QTL of European origin were not overlooked on 5A, markers for two

additional intervals (QFhs.inra-5a2 and QFhs.inra-5a3) identified in Renan by Gervais

et al (2003) were included in the current study. However, once again, there was no

match with the Renan haplotype among test varieties.

A less widely deployed source of resistance associated with the 1RS-1BL rye

translocation in F201R (Shen et al, 2003a) was also included in the current study. Use

of point inoculation by Shen et al. (2003) indicated that the 1B QTL in F201R

conditions significant resistance to spread (type II resistance, sensu Schoeder &

Christensen, 1963). In the current study, a large group of test varieties with varying

FHB reactions had the 1RS-1BL translocation and shared the same allele as F201R at

the Barc008 locus. However, only the German cv Petrus and the reference varieties

Wuhan 2-37E, CM82036, Aurora and WEK0609 had a null allele at the Xgwm018 locus

suggesting that this locus may be diagnostic for this F201R QTL in combination with

the rye translocation and the 191 base pair (bp) allele of Barc008. In contrast to

F201R where the 1B locus is of major effect, Buerstmayr et al (2002) reported that

type II resistance on chromosome 1B in CM82036, associated with the high molecular

weight (HMW) glutenin locus (Glu-B1), was of only minor effect. This may have been

because more potent loci were segregating for resistance in CM82036 compared to

F201R diminishing its apparent effect, however, analysis in WEK0609 indicated that

the 1B QTL is of almost equal importance to that of Qfhs.ndsu-3BS (Gosman et al,

unpublished). Analysis of the protein banding pattern at Glu-B1 locus indicates that

CM82036 has the ‘i’ allele whereas other varieties sharing the F201R haplotype have

the ‘c’ allele (data not shown) indicating that CM82036 may possess a less potent

variant of the QTL. In the current study, the 1RS-1BL translocation is associated with

an approximately 190 bp product at the Barc008 locus in every case. However,

according to the consensus map of Somers et al (2004) Barc008 is on the short arm

of 1B and should, therefore, be null in the presence of the rye translocation. It is

possible that there is a priming site for Barc008 on 1RS in rye however; it is more

likely that this locus is in fact on the long arm, rather than the short arm.

Varieties with the lowest levels of symptom development in the current study

also had toxin levels in grain that were well below proposed EU action limits. Three

varieties from the 2003 UK National List, Soissons, Spark and Vector had a stable

reaction over environments and should, therefore, be suitable for introducing FHB

resistance into UK breeding programs. Studies of combining ability in wheat suggest

37

that high levels of FHB resistance can be achieved from crosses between parental

lines with only moderate levels of resistance (Buerstmayr et al, 1999 and

Buerstamayr et al, 2004). In view of this, genetic analysis is under way to determine

the genomic location and significance of FHB resistance in Soissons and Spark with a

view to eventually ‘pyramiding’ resistance in germplasm with more favorable

agronomic characteristics. SSR haplotyping indicated that these varieties do not

possess any of the FHB resistance QTL found in Sumai-3, Renan or F201R. The

resistance in these varieties, therefore, appears to be novel.

None of the FHB resistant varieties from the UK or mainland Europe had SSR

haplotypes indicating that their resistance is derived from Sumai-3. The introduction

of potent QTL for resistance to FHB from this source should also provide a means to

enhance the FHB resistance of European varieties.

38

4 Identification and mapping of QTL for FHB resistance and other

agronomic traits in different mapping populations

Following initial studies to identify FHB resistance, detailed studies were undertaken to

identify FHB QTL from three resistance sources. Soissons and Spark were found to be

the most resistant UK varieties and QTL analysis was undertaken of Spark x Rialto and

Soissons x Orvantis doubled haploid populations. The third population was of

recombinant inbred lines from RL4137 x Timgalen. Field and polytunnel disease trials

were established and each line scored at several locations. Genetic maps were

produced for all three crosses and QTL analysis carried out for FHB resistance traits

and for selected morphological traits to determine the genetic basis of FHB resistance

and associations with other traits. These the results of these studies have been

discussed in this section.

4.1 Spark/Rialto DH population

4.1.1 Results

4.1.1.1 Trait analysis in the population

Spark, Rialto and the DH lines were phenotyped for FHB across three environments.

The frequency distribution for AUDPC in the ADVANTA2005, NIAB2005 and JIC2006

trials was continuous with transgressive segregation and slightly skewed towards

greater susceptibility, with the population mean being greater than the mid-parent

mean (Table 4.1).

Table 4.1 Summary statistics for fusarium head blight (FHB) disease and plant height assessed in different experiments

Expt. site & trait Spark Rialto MPV PM Range FHB-AUDPC ADVANTA2005 246 557 401.5 529 128-1100 NIAB2005 208 339.5 273.8 294 37-695 JIC2006 144 792.5 468 515 61-1270 Plant height (cm) JIC2006 103.3 101.1 102.2 102 82-130

AUDPC= area under the disease progress curve, MPV = mid-parent value PM = population mean

The analysis of variance (ANOVA) and broad sense heritability (h2) were

calculated for AUDPC in each trial and across environments and for plant height at the

JIC2006 (Table 4.2). Within-trial heritability (h2) ranged from 0.48 to 0.88 and was

highest for AUDPC-dpi at JIC2006. Across environments, the heritability was 0.48.

39

4.1.1.2 Association between height and disease

The parental lines had very similar plant heights: Spark (103.25cm) and Rialto

(101.08cm). The frequency distribution for height within the DH population was

continuous, ranging from 82-130 cm, and followed a normal distribution pattern

centred on the two parents, with a population mean of 101.7cm. Analysis indicated

that most of the differences in height within this population have a genetic basis with

a high heritability (h2=87%), (Table 4.2). A weak/moderate negative correlation was

detected between plant height at J2006 and AUDPC with correlation coefficient 0.39

(P=<0.001), indicating that plant height has some effect on resistance.

Table 4.2 Analysis of variance and heritability of different traits in individual trials and across environments Expt. Site and year Trait Variance factor DF MS

P-value

h2

ADVATA2005 AUDPC Genotypes 121 56045 <.001 0.48 Replicates 1 99487 0.07 Genotype x Replicates 121 30288 0.5 Residual 121 30288 NIAB2005 AUDPC Genotypes 121 39255 <.001 0.62 Replicates 1 18136 0.22 Genotype x Replicate 121 11864 0.5 Residual 121 11864 JIC2006 AUDPC Genotypes 121 194638 <.001 0.88 Replicates 1 442255 <.001 Genotypes x Replicates 121 13626 0.5 Residual 121 13626 Across sited AUDPC Genotypes 121 115644 <.001 0.48 Locations 2 4189900 <.001 Genotypes x Locations 237 96395 <.001 Residual 424 27539 JIC2006 Plant height Genotypes 121 231.97 <.001 0.87 Replicates 1 25.94 0.24 Genotypes x Replicates 121 15.45 0.8 Residual 53 18.67

DF = degrees of freedom, MS = variance expressed as mean squares, h2 = experimental

repeatability/heritability.

4.1.1.3 Identification of FHB resistance QTL

A total of ten FHB resistance QTL exceeding genome wide LOD threshold significance

were detected. Five QTL were consistently detected in more than one environment

(Table 4.3 and Fig. 4.1). Four major QTL, one each on 1B, 4A, 4D and 6A were

detected. LG 3A detected two QTL (Qfhs.jic-3a.1 and Qfhs.jic-3a.2) which are

separated by about 50 cM in more than one environment. The QTL Qfhs.jic-3a.1

closely linked to WMC11(2) and DArT markers, wPt-7992, were contributed by the

alleles from Rialto. Similarly, Qfhs.jic-3a.2 closely linked two SSR markers, GWM497

40

and BAC19, was from the other parent Spark. The two major QTL detected in this

study are of particular interest. Qfhs.jic-4d.2 on 4D with LOD value 6.8 to 28.2 (R2 =

21.4 to 52.7 % except at Elsoms), is a consistent QTL identified in all the

environments. This QTL was closely linked to Rht-D1b (Rht2), a sequence based

marker and PSP3103, an SSR marker. This LG also detected a minor QTL at JIC for

AUDPC-days. Surprisingly the alleles from Spark contributed for both these QTL.

Another major QTL closely linked to the DArT marker, wPt-8833, was detected in

more than one environment. Two minor QTL, one each on initial segment of 6A and

terminal segment of 7A, are only suggestive (Table 4.3 and Fig. 4.1).

wPt-84460.0Xgwm3892.0Xgwm5338.2

Xgwm49319.6wPt-721224.3

stm12tgag39.8Xwmc5442.6stm40tcac55.0stm5.2tgag57.3Xgwm37659.3Xpsp314461.0Xbarc16462.1Xgwm13164.0Xwmc5666.1Xbarc22972.9

wPt-576992.0

3B

R

Xgwm6350.0

wPt-592628.3

wPt-483536.9Xgwm13049.7Xgwm27656.1Xbarc154(2)57.9Xbarc12760.1Xpsp300160.6Xpsp305069.8wPt-197472.6Xwmc975.3Xpsp3050(2)76.7Xbarc2977.6Xbarc19579.4Xbarc2181.1wPt-729982.5Xgwm33291.5wPt-982492.8Xpsp3094(2)94.6Xgwm28295.6wPt-4553109.2wPt-0961110.2Xwmc116114.9wmc273(1)131.7wPt-7947145.7wPt-2501146.7

7A

S

S

3AXwmc11(2)0.0Xbarc120.9wPt-79923.4

wPt-168825.1

wPt-440739.8Xgwm49754.0Xwmc26458.8wPt-921560.5Xbarc1962.0Xgwm262.9Xpsp304763.3Xbarc4566.5

Xgwm15583.6

R

R

S

S

0.0 Xgwm443

wPt-4131Xgwm205

Xbarc14133.7

Xbarc15159.3

wPt-037369.8

5A

10.712.5 S

Xgwm614(2)

Xgwm210(2)

143.0147.0

Xwm311148.4151.0152.0

2A0.0

Xwmc40712.4

25.4Xgwm63630.1stm8acag33.9

Xpsp315347.9

Xgwm35964.0

Xgwm33984.9Xgwm9593.9Xgwm27594.0Xgwm51596.3Xgwm312117.1Xgwm294121.8wPt-1913133.0Xgwm349134.7wPt-6894139.0Xwmc181139.7wPt-9793140.8Xgwm356Xgwm382

wPt-5887Xbarc122

S

R

Xgwm2650.0

Xgwm19219.0Xgwm5521.0Xpsp310324.0wPt-580926.0XRht-D1b29.0wPt-047230.0wPt-071031.0wPt-883638.0

wPt-457262.0

4D

S

SS

SS S

Xbarc1370.0Xgwm112.8Xgwm4135.2Xgwm187.3HMW7+810.3wPt-070511.5Xgwm27312.6wPt-205214.3Xgwm15318.2Xgwm27420.7wPt-903222.0Xgwm26823.8Xwmc4426.9wPt-094442.0Xpsp310052.3Xgwm25960.3

1B

R

Xgwm3340.0wPt-690410.9wPt-068912.2wPt-352413.4wPt-826618.5Xgdm3622.8wPt-883335.4Xbarc2339.4wPt-706342.0Xwmc3244.3Xpsp302947.1Xpsp307148.2Xwmc25650.1Xwmc20152.8Xwmc17954.4Xbarc11358.8Xgwm57067.5

6A

S

S

S

S

Identification key:

: JIC Height

: J2006 AUPDC

: A2005 AUPDC : N2005 AUPDC

: AUPDC across sites

S : Spark, R : Rialto

Fig. 4.1 Linkage maps of chromosome segments constructed from the Spark x Rialto doubled haploid population. Putative QTL positions for FHB resistance and plant height are shown on the right of each linkage group. Genetic distances are shown in cM to the left of each linkage group. Major QTL have been indicated with an arrow and loci in bold are closest to the peak LOD score.

41

4.1.1.4 Plant height QTL

The details on the genomic regions associated with plant height are shown in the

Table 4.3. From both the trials, a total of five QTL which were above the LOD

threshold value 2.9, were detected. They were mapped on LGs 2A, 3A, 3B, 4D and

6A. A major QTL contributed by Spark on LG 4D with closet marker Rht-D1 (LOD

value of about 29 and R2 = 54 %), was detected in both the trials. The LG 3B

detected two minor QTL (about 40 cM apart) contributed by Rialto. Although two QTL

contributed by Spark were mapped on 2B, they were only 10 cM apart and therefore it

could just be single QTL.

42

Table 4.3 Summary of QTL for FHB resistance and plant height identified in Spark/Rialto DH population in different experiments

Name Closest marker Origin

Position ADVANTA2005 NIAB2005 JIC2006 Across Expt.

LOD R2 LOD R2 LOD R2 LOD R2 FHB resistance: Qfhs.jic-1b wPt-0705 Rialto 11.5(1B) _ _ 5.2 13.4 _ _ _ _ Qfhs.jic-2a Xgwm515 Spark 96.3(2A) _ _ 2.9 7 _ _ _ _ Qfhs.jic-3a.1 Xwmc11(2) Rialto 0(3A) _ _ _ _ 2.7’ 2.7 _ _ Qfhs.jic-3a.1 wPt-7992 Rialto 3.4(3A) _ _ _ _ _ _ 2.8 4 Qfhs.jic-3a.2 Xgwm497 Spark 54(3A) _ _ 3.3 8 _ _ _ _ Qfhs.jic-3a.2 Xbarc19 Spark 62(3A) 2.8 7.6 _ _ _ _ _ _ Qfhs.jic-4d.1 Xpsp3103 Spark 24(4D) _ _ 9.3 25.5 _ _ _ - Qfhs.jic-4d.1 XRht-D1b Spark 29(4D) 6.8 21.4 _ _ 28.2 52.7 23.5 50.9 Qfhs.jic-4d.2 Xgwm265 Spark 0(4D) _ _ _ _ 3.1 3.1 _ _ Qfhs.jic-5a Xgwm443 Spark 0(5A) _ _ _ _ 3.5 3.9 _ _ Qfhs.jic-6a.1 Xgwm334 Spark 0(6A) 2.2’ 6.2 _ _ _ _ _ _ Qfhs.jic-6a.2 wPt-8833 Spark 35.4(6A) _ _ _ _ 8.3 10.5 5.2 7.8 Qfhs.jic-7a Xpsp3050(2) Spark 76.7(7A) _ _ _ _ 2.9 3.7 _ _ Qfhs.jic-7a wPt-7299 Spark 82.5(7A) _ _ _ _ 2.2’ 5.2 _ _ Plant height: Xgwm356 Rialto 143(2A) _ _ _ _ 6.9 8 _ _ wPt-7212 Rialto 24.3(3B) _ _ _ _ 3.3 3.7 _ _ XRht-D1b Spark 29(4D) _ _ _ _ 29.2 53.1 - _ Xpsp3071 Spark 48.2(6A) _ _ _ _ 2.9 3.1 _ _

R2 = percentage phenotypic variance explained, ’= LOD below the permutation test threshold for significance, _ = no data

43

4.2 Soissons/Orvantis DH population

4.2.1 Results

4.2.1.1 Fusarium head blight resistance

FHB severity showed a continuous distribution both in individual environments and for

mean over environments (data not shown). The distributions were slightly skewed

towards resistant parent Soissons. The segregation of FHB severities in the population

differed between environments. Transgressive segregation was noticed in all

environments and was more towards Soissons in JIC2006 and NIAB2005 while it was

more towards susceptible parent Oravantis in CSL2005. The parents displayed

different FHB severities in individual environment (Table 4.4). The parent means (mid

parent value) were greater than the population means except in NIAB2005 (Table 1).

The plant height (PH) showed normal distribution with nearly equal number of

trangressive segregants on either side of the parents (data not shown). The PH of

Soissons and Orvantis were 90.3 cm and 96.1 cm, respectively (Table 4.4).

Table 4.4. Summary statistics for plant height and AUDPC in Soissons/Orvantis DH

population

Expt. site Trait Soisson Orvantis MPV Range PM

CSL2005 AUDPC 0 519.2 259.5 743.2 238.7

JIC2006 AUDPC 208.6 1838.4 1023.5 1670 470.6

NIAB2005 Incidence 6.6 45 25.8 95 34.6

Severity 4.1 13.9 9 58.8 13.7

AUDPC 34 677 356 5508 659

Across sites AUDPC 80.9 1011 556 2382 546

JIC2006 Plant height 90.3 96.1 93.2 70 92.9

MPV = mid-parent value, PM = population mean

4.2.1.2 Correlations, ANOVA and broad sense heritability

Correlations of AUDPC from all environments was found to be significant (P <0.001)

(data not shown) and the correlation coefficient (R2) ranged from 0.138 to 0.839 with

an average R2 value 0.443. PH also showed weak to moderate levels of correlation

with AUDPC (R2 = 0.04 to 0.138). The genotypes were found to be highly significant in

all the environments (P= 0.007 to <0.001) and moderate significance of replication

44

was also noticed in CSL2005 and NIAB2005 (Table 4.5). The broad sense heritability

(h2) ranged from 42 to 90 % with an average heritability of 69 %.

4.2.1.3 FHB and Plant height QTL

The map based QTL analysis revealed the association of genomic regions on 1B, 3B,

4B, 4D and 7A and three of these QTL were found to be significant (Table 4.6 and Fig.

4.2). The alleles from Soissons contributed for all these QTL except for the QTL on 1B.

A major QTL Qfhs.jic.4d was detected in all the environments. Two minor QTL

Qfhs.jic.3b and Qfhs.jic.7a were observed in N2006.

A total of five QTL on three wheat chromosomes namely, 1BL, 4BL, 4DS and

7AL were found to be associated with PH and the alleles from both parents contributed

for the QTL. Of these, QTL on 4B and 4D with the closest loci Rht-B1b and Rht-D1b,

respectively were found to be major PH-QTL in this population. The alleles from

Soissons and Orvantis contributed for the QTL on 4DS and 4BL, respectively. Two

minor QTL one each from Soissons and Orvantis were mapped on 1BS and 1BL. As

majority of the makers on the map were DArTs, either consensus wheat maps

published on the Triticarte website or wheat genetic maps (DArT markers anchored

with SSRs) was used to identify the positions of QTL mapped in this study.

A major FHB QTL Qfhs.jic.4d co-localised with a major PH-QTL with the closest

marker Rht-D1b on 4DS and the alleles from Soissons contributed for both these QTL.

Similarly, a minor FHB QTL contributed by the alleles from Orvantis also co-localised

with major PH-QTL with the closest maker Rht-B1b on 4BS. None of the other PH-QTL

were co-localized with other FHB-QTL detected in this study.

4.2.1.4 Phenotyping of Rht-B1b and Rht-D1b near-isogenic lines

Experiment 1: In the Poly-tunnel, when Mercia and Maris Huntsman were spray

inoculated with F. culmorum or F. graminearum, both variety (P<0.001, Table 4) and

Rht status (P=0.002) were highly significant, no interaction of Rht by variety

(P=0.451) was observed. Compared to Rht (tall), both Rht-B1b (P = 0.090) and Rht-

D1b (P = 0.087) were moderately significant, however no significant differences (P =

0.987) existed between them. The mean AUDPC were 11209, 11784 and 6861 for

Rht-B1b, Rht-D1b and Rht (tall), respectively.

When the same genotypes were sprayed with F. culmorum, the Rht status was

found to be highly significant (P = 0.010, Table 4.7) and the Rht by variety interaction

showed moderate effect (P = 0.237), variety was not significant (Table 4.7). The

mean AUDPC of Rht-D1b (13417) was significantly higher (P<0.001) than Rht (tall).

45

No significant differences existed between Rht-D-B1b (P = 0.109) and Rht (tall), and

Rht-B1b and Rht-D1b (P = 0.180).

46

Table 4.5 Variance components identified using generalised linear model for FHB-AUDPC and plant height in the Soissons/Orvantis

DH population

Source of variation Disease Plant height NIAB2005 CSL2005 JIC2006 Across sites

M.S. P-

value M.S. P-

value M.S. P-value M.S. P-

value M.S. P-

value

Genotype (DH line) 3.77E+08 0.007 1171490 <.001 486578 <.001 631795 <.001 137.962 <.001 Genotype x Replication 2.85E+07 0.5 432762 0.5 27577 0.5 431372 <.001 8.384 0.5 Replication or Expt. Site 1.28E+08 0.082 3884302 0.003 2529 0.762 35521978 0.486 106.381 <.001 Residual error 2.40E+07 432762 27577 430715 8.384 R or h2 0.88 0.58 0.9 0.42 0.9

M.S. refers to mean square, R refers to experimental repeatability and h2 refers to broad sense heritability.

47

Table 4.6 Summary of QTL for FHB resistance identified in Soissons/Orvantis DH population in N2005, C2005 and J2006 trials

FHB trait Name Closest Parent Position JIC2006 NIAB2005 CSL2005 Pooled data

locus LOD R2 LOD R2 LOD R2 LOD R2

AUDPC Qfhs.jic.4b RhtB-1b Soissons 0(4BS) 2.1$ 5.3 _ _ 1.6$ 4.2 AUDPC Qfhs.jic.4d.1 wPt-7569 Soissons 0(4DS) 6.4 16.1 _ _ 4 10.6 7.5 18.4 Incidence Qfhs.jic.1b wPt-1403 Orvantis 3.6(1BL) _ _ 2.3$ 5.4 _ _ _ _ Qfhs.jic.7a wPt-7034 Soissons 4.4(7AL) _ _ 3.4 8.5 _ _ _ _ Qfhs.jic.4d.2 RhtD-1b Soissons 9.3(4DS) _ _ 3.2 7.4 _ _ _ _ Severity Qfhs.jic.3b wPt-2559 Soissons 3.9(3BL) _ _ 2.5 6.1 _ _ _ _ Qfhs.jic.4d.1 wPt-0710 Soissons 0(4DS) _ _ 2.6 6.1 _ _ _ _ Disease Index Qfhs.jic.3b wPt-2559 Soissons 3.9(3BL) _ _ 2.6 6.7 _ _ _ _ Qfhs.jic.4d.1 wPt-0710 Soissons 0(4DS) _ _ 2.8 7.1 _ _ _ _ Qfhs.jic.7a wPt-6447 Soissons 0(7AL) _ _ 2.2$ 5.3 _ _ _ _ Plant height wPt-5562 Soissons 0(1B) 3 5.1 _ _ _ _ _ _ wPt-9809 Orvantis 14.7(1B) 4.1 7.4 _ _ _ _ _ _ Rht-B1b Orvantis 0(4B) 8.6 15.9 _ _ _ _ _ _ Rht-D1b Soissons 9.3(4D) 11.6 21.9 _ _ _ _ _ _ wPt-2780 Orvantis 0(7A) 3.2 5.4 _ _ _ _ _ _

LOD = Log of adds ratio, R2 = percentage variance explained and $ = below LOD threshold

48

Identification key:

: Plant Height

: AUDPC-N2005

: AUDPC-J2006

: AUDPC-C2005

: AUDPC-mean over experiments

wPt-52650.0wPt-82920.1

wPt-360814.6

4B-2

Rht-B1b0.0wPt-75692.9wPt-17084.8wPt-39916.7wPt-53347.7

wPt-060821.6

4B-1

O

1B-2

wPt-69750.0wPt-14033.6

wPt-980914.7 O

S1B-1

wPt-29880.0wPt-15603.5wPt-86275.7wPt-22306.1wPt-39278.6wPt-346515.0

wPt-257525.2 wPt-580127.1wPt-375329.3wPt-897137.6 wPt-0974wPt-272539.5wPt-191139.6

wPt-556247.0

S

7A-1

wPt-87890.0wPt-84184.6wPt-52575.4wPt-603411.3wPt-151012.5

wPt-703429.5wPt-965130.9wPt-483533.4wPt-644733.9

wPt-27800.0wPt-5003wPt-22915.0

O7A-2

3B-1

wPt-13360.0wPt-17410.8wPt-24641.1wPt-69651.9

3B-3 S

wPt-25590.0wPt-50721.5wPt-75263.9

wPt-84460.0wPt-72250.9

wPt-186712.0wPt-360915.3

wPt-353627.6wPt-030227.7wPt-303830.9wPt-721232.3

wPt-583658.3wPt-077360.2wPt-621667.9wPt-957968.1wPt-148468.6

wPt-9912104.7

3B-2 4DS S S SS

wPt-88360.0

Rht-D1b24.0

wPt-047232.0wPt-071033.2

Figure 4.2 Linkage maps of Soissons/Orvantis DH population with QTL positions for FHB resistance and plant height. Genetic distances are shown in cM to the left of each linkage map. S and O refer to alleles contributed by Soissons and Orvantis, respectively. Only significant QTL have been indicated in the figure

In the point inoculation, similar to previous experiments, the Rht status was highly

significant and the variety was also significant but no Rht by variety interaction was

observed. Both Rht-B1b and Rht tall (P = 0.001, Table 4.7) and Rht-B1b and Rht-D1b

(P<0.001) were found to be highly significantly different. The mean AUDPC of Rht-B1b

(6.492679) was less than the Rht (tall) (8.194712) and Rht-D1b (8.784821).

49

Table 4.7 Analysis of variance and t-test results of FHB disease levels on rht, Rht-B1b and Rht-D1b near isogenic lines of winter wheat varieties Mercia and Maris Huntsman in a field trial and following spray and point inoculation in a polytunnel Experiment Analysis of Variance T-test

Source of variation d.f. m.s. v.r. F.pr. Comparison F.pr.

Field trial Variety 1 1680556 0.29 0.6 Rht-B1b with rht 0.109 Mercia and Huntsman Rht status 2 4016667 6.95 0.01 Rht-D1b with rht <0.001

Rht x Variety 2 938889 1.62 0.237 Between Rht-B1b and Rht-D1b 0.18

Error 12 5777778 Spray inoculation Rht status 2 108042823 4.08 0.019 Rht-B1b with rht 0.015 Mercia Replicate 3 163413722 6.18 <0.001 Rht-D1b with rht 0.013

Rht status x Block 6 47578640 1.8 0.104

Between Rht-B1b and Rht-D1b 0.853

Error 134 26457077 Spray inoculation Rht status 2 575215 5.48 0.007 Rht-B1b with rht 0.005 Huntsman Replicate 2 55077 0.52 0.595 Rht-D1b with rht 0.005

Rht status x Block 4 97709 0.93 0.453

Between Rht-B1b and Rht-D1b 0.573

Error 54 105030 Point inoculation Variety 1 49.77 7.45 0.007 Rht-B1b with rht 0.001 Mercia and Huntsman Rht status 2 73.96 11.07 <0.001 Rht-D1b with rht 0.269

Rht x Variety 2 5.17 0.77 0.463 Between Rht-B1b and Rht-D1b 0.001

Error 154 6.68 d.f., refers to degrees of freedom; m.s., refers to mean square, v.r., refers to variance ratio and F.pr., refers to F-probability.

50

4.3 RL4137/Timgalen Recombinant Inbred population

4.3.1 Results

4.3.1.1 Performance of RILs

The PH showed normal distribution and the parents RL4137 and Timgalen 97.97 and

162.33 cm, respectively and the population mean 121.1 cm (Table 4.8). A Few

trangressive segregants towards RL4137 were evident. The population also showed

similar trend for awns. Correlations between field and pot experiments for AUDPC was

highly significant (P<0.01) and the correlation coefficient ranged from 0.5 (between

JIC2000 and NIAB2006) to 0.9 (JIC1999 and JIC2000) (Table 4.9). The PH, awns and

SW also showed a continuous distribution. The PH showed negative relationship with

AUDPC and the correlation coefficient ranged -0.22 (P = 173 with JIC1999) to -0.47

(P<0.001 with NIAB2006). The awns also showed weak but non-significant negative

relationship (r = -0.13 to -0.23) except with JIC2000 (r = 0.12, P = 0.062) (Table

4.9). Similarly, the spikelet weight showed significant negative relationship and the

correlation coefficient ranged from -0.61 to -0.80 (P = 0.001 to <0.001). For AUDPC,

the genotypes were always highly significant in all the three experiments (P<0.001)

and blocks were not significant except in JIC2000 (P<0.001) (Table 4.10). However,

significant block and block by genotype interactions were observed for PH, awns and

WIS. The repeatability ranged from 0.68 (JIC1999) to 0.71 (JIC2000) and 0.58 (WIS)

to 0.9 (PH) for AUDPC and FHB related traits, respectively (Table 4.10).

Table 4.8 Summary statistics for Fusarium head blight (FHB) and associated traits

disease assessed in different trails

Experiment and trait recorded

RL4137 Timgalen MPV PM Range

JIC1999-AUDPC 545.64 1161.08 853.36 835.9 50-1796.7

JIC2000-AUDPC 357.38 983.85 670.62 477.4 0-1540

NIAB2006-AUDPC 75.31 435.88 255.59 276.7 12-835

JIC2000-Spkilet weight 0.08 0.04 0.06 0.59 0.04-1.49 JIC2000-Awns 1 5 2.5 4.336 1-5

JIC2000-Plant height 162.3 97.7 130.0 121.1 73-170 AUDPC = area under disease progress curve, MPV = mid-parent value, PM = population mean Awns were measured on 1 to 5 scale.

51

Table 4.9 Correlation coefficients of AUDPC and other FHB related traits in

RL4137/Timgalen RIL population

Trait J1999 AUDPC

J2000 AUDPC

N2006 AUDPC

Pooled AUDPC

J2000 PH

J2000 A

J2000 WIS

J1999-AUDPC 1

J2000-AUDPC 0.70 1

N2006-AUDPC 0.56 0.51 1

Pooled-AUDPC 0.90 0.88 0.77 1

J2000-PH -0.22 -0.42 -0.47 -0.42 1

J2000-A -0.22

0.116 -0.27 -0.13 -0.46 1

J2000-WIS -0.68 -0.75 -0.61 -0.80 0.54 -0.06 1 J = JIC; N = NIAB; PH = plant height; A = awns; WIS = weight of infected spikelet

Table 4.10 Components of variation for AUDPC and other FHB associated traits in

RL4137/Timgalen RIL population

Source of FHB-AUDPC FHB associated traits

variation m.s. v.r. F pr. m.s. v.r. F pr.

JIC1999 Plant height-JIC2000

Blocks 239953 2.66 0.035 1402.96 36.28 <.001

Genotypes 430371 4.77 <.001 3347.97 86.57 <.001

Residual 90194 38.67

h2 0.7 0.94

JIC2000 Awns-JIC2000

Blocks 819205 10.08 <.001 0.2808 1.53 0.218

Genotype 458747 5.65 <.001 7.2132 39.25 <.001

Residual 81256 0.1838

h2 0.71 0.92

NIAB2006 Spikelet weight-J2000

Blocks 143 0.01 0.934 0.00224 4.48 0.012

Genotypes 54885 2.69 0.007 0.00457 9.14 <.001

Residual 20396 0.0005

h2 0.68 0.58 m.s. = mean square, v.r. = variance ratio, Fpr. = F probability, h2 = repeatability, J = JIC, N = NIAB

4.3.1.2 Mapping of QTL for FHB resistance and associated traits

The genetic map based QTL analysis was performed in single experiment and across

experiments on AUDPC values. The pooled average AUDPC from the three

experiments was treated as another environment. The analysis identified a total of six

52

putative QTL for AUDPC on the chromosomes 1B, 2A, 2B, 3A, 6B and 7A (Table 4.11

and Fig. 4.3). The alleles from RL4137 contributed for all the QTL except for the QTL

on 6B. Among these, three major QTL detected one each on chromosomes 2B, 3A and

6B were detected. Three QTL namely Qfhs.jic-1b, Qfhs.jic-2b and Qfhs.jic-6b were

detected in more than one environment. The alleles for the QTL on 1B and 3B were

contributed by RL4137 and that of QTL on 6B was contributed by the alleles from

Timgalen. Both PH and awns detected a QTL on 2B with the closet AFLP marker

S13/M23G. The alleles from RL4137 and Timgalen, respectively contributed for PH and

awns QTL. SW detected one QTL each on 2B and 6A and the alleles from RL4137

contributed for two SW QTL one each on 2B and 6A. The QTL for PH, WIS and awns

co-localized with a major FHB resistance QTL Qfhs.jic-2b. The alleles from RL4137

contributed for the FHB, PH and WIS QTL and the QTL for awns was contributed by

the alleles from Timgalen.

53

Table 4.11 Summary of QTL identified for FHB resistance and plant height in RL4137/Timgalen RILs in different experiments

Trait QTL Closest Position Origin JIC1999 JIC2000 NIAB2006 Over expt. marker LOD R2 LOD R2 LOD R2 LOD R2 AUDPC Qfhs.jic-1b wPt-6425 51.3(1B) RL4137 _ _ _ _ 2.6 ' 9.9 2.4 ' 9.1 Qfhs.jic-2b wPt-5292 4.6(2B) RL4137 _ _ _ _ 4.8 21.5 _ _ Qfhs.jic-2b S24/M16I 6.2(2B) RL4137 _ _ 5.2 19.8 _ _ 2.1' 8.4 Qfhs.jic-3a S22/M14F 48.1(3A) RL4137 _ _ _ _ 4.6 20.8 _ _ Qfhs.jic-6a wPt-9132 12.3(6A) RL4137 3.4 25.8 _ _ _ _ _ _ Qfhs.jic-6b wPt-3376 16.6(6B) Timgalen _ _ _ _ _ _ 2.4' 9.2 Qfhs.jic-6b wPt-4930 17.3(6B) Timgalen _ _ 3.8 14 _ _ _ _ Qfhs.jic-7a wPt-3836 3.1(7A) RL4137 _ _ 2.1' 7.4 _ _ _ _ WIS Qfhs.jic-2b S13/M23G 5.5(2B) RL4137 _ _ _ _ 3.5 33.9 _ _ Qfhs.jic-6a wPt-9132 12.3(6A) RL4137 _ _ _ _ 4.2 40.4 _ _ DON-AUG Qfhs.jic-2b S24/M16I 6.2(2B) RL4137 2.1' 23.3 _ _ _ _ _ _ Qfhs.jic-7A wPt-6273 3.3(7A) RL4137 2.3 ' 20.3 _ _ _ _ _ _ Plant height S13/M23G 5.5(2B) RL4137 _ _ 15.9 47.9 _ _ _ _ wPt-7524 3.4(4A) Timgalen _ _ 2.9 5.4 _ _ _ _ wPt-9454 57.8(5B) RL4137 _ _ 2.7 ' 5.9 _ _ _ _ Awns S13/M23G 5.5(2B) Timgalen _ _ 3.4 14.6 _ _ _ _

R2 = Percentage phenotypic variance explained, ' = LOD below the permutation test threshold for significance WIS = weight of infected spikelet, AUDPC = area under disease progress curve, DON = deoxynivalenol, AUG = area under growth

54

S21/M13K0.0S16/M12K2.3wPt-30603.8wPt-96424.7S12/M14A7.4S20/M22P10.6wPt-492413.0S20/M22G16.0

wPt-104824.1

6B-2

S16/M12G0.0

S14/M20F6.4wPt-01717.2wPt-61168.3wPt-30458.4S20/M22L9.6

6B-3

S20/M25C0.0

wPt-31167.9wPt-52348.4

wPt-209516.2wPt-456416.5wPt-337616.6wPt-493017.3

6B-1T T

wPt-01000.0GWM5010.3wPt-48231.4S13/M23B1.9wPt-00793.0wPt-5242wPt-93503.6

GWM4104.1wPt-52924.6S13/M23GS20/M25A

5.5

S24/M16I6.2

S20/M19B7.5

GWM4298.7wPt-61588.9wPt-7995wPt-3565

9.3

wPt-670610.4wPt-631110.5wPt-823511.6wPt-622311.9wPt-740821.4wPt-832625.0 S12/M14E

26.1 S24/M16JS21/M13U29.1S21/M13G29.4wPt-242529.5 wPt-0049wPt-4559wPt-3378wPt-2135

29.6

wPt-537433.4

2B

RR

R R R TR

:

:

:

AUDPC-J1999

AUDPC-J2000

AUDPC-N2006

Identification key:

: Pooled AUDPC

WIS-J2000:

: PH-J2000

: Awnedness-J2000

: DON-AUG

Figure 4.3 Putative QTL positions identified in RL4137/Timgalen for the major consistent FHB resistance along with the co-localized QTL for the DON-tolerance (effect of DON on seed germination and growth, DON-AUG), plant height (PH), weight of infected spikelet (WIS) and awns are shown on the right of each linkage map. J and N refer to JIC and NIAB trail sites, respectively. Genetic distances are shown in cM to the left of each linkage map. R and T refer to alleles contributed by RL4137 and Timgalen, respectively. 4.4 Discussion

4.4.1 FHB resistance QTL in Spark and the role of Rht-D1b semi-

dwarfing allele in FHB susceptibility

These studies have clearly demonstrated that the Rht-D1b semi-dwarfing allele is

associated with increased susceptibility to FHB. Furthermore, we have shown that the

effect relates to a decrease in Type I resistance with no apparent effect on Type II

resistance. Numerous authors have reported an association between plant height and

resistance to FHB (Mesterhazy, 1995; Somers et al, 2003; Klahr et al, 2007). It has

been proposed that in natural FHB epidemics, the smaller distance between the soil

and the first leaf and between the flag leaf and the head in short genotypes may

55

result in greater exposure to primary inoculum where it is splash dispersed from the

soil or stem-base as suggested by (Jenkinson & Parry, 1994). Even with direct spray

and the application of mist irrigation the influence of plant height on susceptibility has

been observed (Buerstmayr, et al, 2000; Gervais et al, 2003) supporting the

hypothesis that the microclimate about the head is more conducive to infection in

shorter varieties (Klahr et al, 2007). The importance of the negative correlation

between FHB and plant height in field trials has been found to vary in magnitude,

depending upon the population under study. Somers et al. (2003) observed a very

significant negative correlation (r=-0.65) in progeny from a cross between Wuhan-1

and Maringa whereas Klahr et al (2007) observed an effect in only two of four

environments (r=-0.27 and -0.32).

Coincident QTL for FHB resistance traits and plant height have been reported in

several studies (Gervais et al, 2003; Somers et al, 2003; Paillard et al, 2004; Steiner

et al, 2004). It has been suggested that taller lines may escape from infection by

having heads further from the soil and that the microclimate about the ear may differ

in short and tall varieties (Somers et al, 2003). However, several studies indicate that

the relationship between FHB resistance and plant height may be more complex.

Somers et al (2003) determined that a QTL on 2DS influencing the accumulation of

DON in grain was coincident with a QTL for plant height although this QTL had no

apparent effect on disease symptoms. However, a second QTL for DON accumulation

(5A) and a QTL for FHB symptoms (4B) were not associated with PH. In a cross

between Renan and Recital QTL for FHB and PH were coincident on 5A whereas the PH

QTL on 4A was not associated with differences in resistance to FHB (Gervais et al,

2003). Steiner et al (2004) found coincident QTL for FHB and PH in a similar position

on 5A derived from Frontana. Paillard and coworkers also found that while some PH

QTL overlapped with those for FHB (e.g. 5B) the main PH QTL (2AL and 5AL) in their

population from a cross between Arina and Forno were not coincident with FHB QTL

(Paillard et al, 2004). A FHB and PH QTL on 6A was found to be coincident in a cross

between Dream and Lynx, although other FHB and PH QTL were independent in this

cross (Schmolke et al, 2005). These authors concluded that the coincidence of FHB

and PH QTL following spray inoculation suggested that it has a genetic basis- either

linkage or pleiotropy rather than being due to escape. While PH may be correlated

with resistance to FHB, the finding that some QTL for PH coincide with those for FHB,

while others do not, supports this view.

The negative effect of the semi-dwarfing alleles Rht-B1b and Rht-D1b (formerly

termed Rht1 and Rht2, respectively) on resistance to FHB has been observed or

56

inferred in previous studies. Hilton et al (1999) observed that, in crosses between

varieties carrying these alleles (semi-dwarf) with those carrying the wild-type alleles

(tall) there was a clear tendency for tall strawed lines to appear more resistant than

short strawed lines. These authors also spray-inoculated near-isogenic lines differing

in Rht-B1 and Rht-D1 alleles with a mixture of Fusarium species and M. nivale. In

most cases lines carrying either the Rht-B1b or Rht-D1b alleles were more susceptible

than lines carrying the wild-type alleles (Hilton et al, 1999). Relative humidity at ear

height was not found to differ between tall and short isogenic lines and it was

concluded that the effect on FHB susceptibility was not due to higher humidity at ear

height in the shorter genotypes.

The negative effect associated with the Rht-D1b allele carried by Rialto has also

been observed in Riband, another UK variety (Draeger et al, 2007). As in the present

study Draeger and colleagues (2007) showed that the main FHB QTL segregating

among progeny from a cross between Arina and Riband co-localised with the Rht-D1

locus. These authors concluded that the effect is not due to plant height per se but

rather to plieotropy or linkage. It was also noted that the great majority of UK wheat

varieties carry the Rht-D1b allele and that they are generally highly susceptible to FHB

(Gosman et al, 2007). In combination with the present work these studies indicate

that Rht-D1b (Rht2) is associated with reduced resistance to FHB but less is known of

the effect of Rht-B1b (Rht1) on resistance. Hilton et al (1999) found no evidence for

differences in the effect of the Rht-B1b and Rht-D1b alleles. However, in a separate

study of a population derived from a cross between Frontana and Remus a FHB QTL

identified on 4B coincident with a PH QTL thought to correspond to Rht-B1b accounted

for only 7.4% of phenotypic variance (Steiner et al, 2004). Further study is required

to determine whether the Rht-B1b and Rht-D1b alleles differ in their influence on

susceptibility to FHB. It is perhaps significant that the semi-dwarfing alleles of the

Rht-B1 and Rht-D1 alleles carried on separate chromosomes (4B and 4D,

respectively) are both associated with increased susceptibility to FHB, suggesting that

the effect may be due to pleiotropy rather than linkage to deleterious genes.

The results from different types of inoculation procedure in the present study

clearly demonstrate that the negative effect of the Rht-D1b allele on FHB resistance

acts on resistance to initial infection (Type 1 resistance) (Schroeder & Christensen,

1963) with little, if any effect on spread within the head (Type II resistance). Other

studies also found that the relationship between resistance to FHB and PH was evident

in field trials measuring Type I resistance (incidence) or a combination of Type I and

Type II resistance (severity) but not where point inoculation was used or spread

57

within the head was measured (Somers et al, 2003; Steiner et al, 2004). It remains to

be determined how the Rht-B1b and Rht-D1b alleles might alter susceptibility to

infection (Type I resistance) and this is the subject of ongoing research.

In addition to the Qfhs.jic-4d.1 coincident QTL for FHB resistance and PH, two

other chromosomes carried QTL for both FHB resistance and PH. On 2A, however, the

allele from Spark contributed the FHB resistance while that for PH was contributed by

Rialto. Additionally, although, QTL for FHB and PH were located on 6A, they did not

overlap.

Two major QTL conferring resistance to FHB were detected on 4D and 6A

insinuating regions for candidate genes/QTL for novel sources for FHB. Our study also

provides molecular markers for MAS and gives access to a previously uncharacterized

source of FHB resistance. Hitherto, the successful application of MAS needs QTL

validation in other genetic backgrounds and germplasms.

In the present study, ten putative FHB QTL were observed to segregate within

the Spark x Rialto population. These were on chromosomes 1B, 2A, 3A, 4D, 5A, 6A

and 7A. The total variance explained for AUDPC by all the QTL across environments

varied being greatest at JIC (76.6%). A total of three major (R2≥10%) and seven

minor (R2 <10%) QTL were mapped. Of these, the alleles from Spark and Rialto

contributed resistance for eight and two QTL, respectively. The contribution to FHB

resistance of alleles from both the parents is not uncommon and previous studies of

FHB resistance in winter wheat have indicated that resistance is controlled by several

loci on different chromosomes, each with only a weak or moderate effect (Paillard et

al, 2004; Schmolke et al, 2005; Draeger et al, 2007). This is in contrast to spring

wheat where several FHB resistance genes of major effect have been identified

(Cuthbert et al, 2006; Cuthbert et al, 2007). An exception appears to be Qfhs.jic4.1

associated with the Rht-D1 locus in winter wheat which explained up to 53% of

phenotypic variance in the Spark x Rialto population used here and up to 24% in the

Arina x Riband population (Draeger et al, 2007).

Other than Qfhs.jic-4d.1, two relatively stable QTL of moderate effect were also

observed to segregate in the Spark x Rialto population. Qfhs.jic-6a.2 on 6A (R2 = 7.8

to 10.5%) and Qfhs.jic-3a.2 on 3A (R2 = 7.6 to 8%), both originating from Spark.

Qfhs.jic-6a.2 appears to be in a similar location to a QTL identified in the FHB

resistant winter wheat variety Dream (Schmolke et al, 2005). These authors reported

that the FHB QTL from Dream partially overlapped with a PH QTL and, interestingly,

Spark also possessed a PH QTL near Qfhs.jic-6a.2 but distinct from this locus. The

Qfhs.jic-3a.2 from Spark is located in a similar position on chromosome 3A to FHB

58

QTL reported from the spring wheat variety Frontana (Steiner et al, 2004) and

Fundulea 201R, a winter wheat line (Shen et al, 2003). Further work is required to

establish the relationship between these three FHB QTL.

Rialto carries the 1BL/1RS wheat-rye translocation (Snape et al, 2007). Several

studies have found that this translocation is associated with Type II resistance to FHB

(Buerstmayr et al, 2002; Shen et al, 2003; Schmolke et al, 2005). In the present

study a QTL of major effect associated with chromosome 1B from Rialto was observed

at a single location (NIAB2005). Whereas the QTL in the CM-82036-derived population

was also only detected in some trials (Buerstmayr et al, 2002) that from Fundulea

201R was consistently expressed across environments (Shen et al, 2003). It may be

that expression of the FHB QTL associated with the 1BL/1RS wheat-rye translocation

is influenced by the genetic background in which it is present.

Spark has previously been shown to be one of the most resistant UK winter

wheat varieties (Gosman et al, 2007). The present work has demonstrated that most

of this effect may be due to the presence of the wild-type (tall) allele at the Rht-D1

locus in this variety in contrast to the semi-dwarfing allele (Rht-D1b) that is present

on most UK varieties. The Rht-D1b allele appears to confer enhanced susceptibility to

FHB by reducing the level of Type I resistance. This allele however, has little or no

effect on Type II resistance. Additional FHB QTL of lesser effect are also present in

Spark (3A and 6A) but additional work is required to establish their value in breeding

programmes.

4.4.2 FHB resistance QTL identified in Soissons and the role of Rht-B1b

and Rht-D1b semi-dwarfing alleles in resistance

The DH population of Soissons (Rht-D1b)/Orvantis (Rht-B1b) was phenotyped for FHB

resistance in the field at NIAB2005, CSL2005 and JIC2006. The genotypes were

always significant both in individual environments and mean over environments. FHB

traits showed moderate to high levels of broad sense heritability. A total of five QTL

regions on 1BL, 3BL, 4BS, 4DS and 7AL appear to be involved in increased resistance

to FHB. The alleles from Soissons contributed for all the QTL except for a minor QTL

on 3BL. Despite variation in the levels of correlation for AUDPC between

environments, a major QTL on 4DS was consistently detected in all the environments

with an explained variance of 6.1 to 18.4 %. In addition to this, two minor QTL one

each on 3BL and 7AL were also detected. Similar to FHB trait, the PH also detected

four QTL and the loci Rht-B1b on 4B and Rht-D1b on 4D were the major height

affecting QTL in this study.

59

In wheat-Fusarium interaction, because of passive resistance, often plants either

completely escape from infection or show fewer FHB symptoms. Thus resistance

breeding may be hindered by the association between FHB resistance with other

important agronomic traits such as the presence or absence of awns, flowering time

and plant height. It is therefore, important to consider pleiotropic effects of such traits

on FHB infection. Generally taller and early heading genotypes show fewer FHB

symptoms (Miedaner, 1997; Hilton et al, 1999; Steiner et al, 2004) and significant

negative association between PH and FHB resistance and/associated traits such as

accumulation of DON in grains has also been reported (Mesterhazy, 1995; Somers et

al, 2003; Draeger et al, 2007; Klahr et al, 2007 and Srinivasachary et al, 2008).

These studies are often supported by the co-localization of QTL for these traits

(Gervais et al, 2003; Paillard et al, 2004; Steiner et al, 2004; Schmolke et al, 2005;

Draeger et al, 2007; McCartney et al, 2007 and Srinivasachary et al, 2008). Similar to

the findings in Arina by Draeger et al (2007) in Arina/Riband cross and Srinivasachary

et al (2008) in Rialto in Spark/Rialto cross containing Rht-D1b, in the current study

also a major PH-QTL co-localized with a major FHB-QTL on 4DS. In addition to this,

Rht-B1b also co-localized with a minor effect FHB QTL on 4B. The long arm of 7A

comprised of a major and a minor LG. A minor PH QTL contributed by the alleles from

Soissons and a PH-QTL contributed by Orvantis was mapped on to two independent

LGs.

Rht-D1b (Rht2), as in previous studies by Hilton et al (1999), Draeger et al

(2007) and Srinivasachary et al (2008) using the isogenic lines of Rht-D1 have

showed the detrimental effect on FHB resistance and but this effect is due to more

than just height, shown by comparison with second semi-dwarfing allele. When a

series of experiments on isogenic lines of Rht-B1 and Rht-D1 by spray (Type I and

Type II) and point inoculations (Type II), it was clear that both alleles compromise for

Type I resistance but Rht-B1b differs from Rht-D1b by showing positive effect for Type

II resistance.

In a cross between 98B69-L47 (carries Rht-B1b) and HC374, McCartney et al

(2007) observed a strong association of a major FHB QTL with plant height as FHB

QTL was in the region of Rht-B1 locus. Wuhan-1, which as responsible for 4B

resistance was associated with increase in plant height. However they could not study

further if this is due is linkage of Rht-B1 and 4B FHB resistance QTL or pleiotrophic

effect of Rht1-B1 on FHB resistance. Similar to our studies, Jia et al (2006) along with

Lin et al (2006) and Lie et al (2007) have also reported a major effect QTL on 4B from

different spring wheats.

60

Some of the QTL detected in this population have also been reported by several

workers in different populations. The studies involving either Sumai 3 or its

derivatives have reported a major QTL on 3BS (Anderson et al, 2001; Bai et al, 1999;

Zhou et al, 2002; Buerstmayr et al, 2002; Somers et al, 2003; Jia et al, 2005; Mardi

et al, 2005; Shen et al, 2002; Ma et al, 2006). Lie et al (2007) have also reported a

FHB-QTL in the centromeric region on 3B. In this study, a minor FHB-QTL was

mapped to the long arm of 3B. Paillard et al (2004) have also mapped a minor QTL on

3BL in Arina (resistant) / Forno (susceptible). Similarly, Gervais et al (2003) also

reported a minor FHB QTL on 3BL. More recently, Klahr and Zimmermann (2007)

have also reported QTL on 3B in winter wheat Cansas, moderately resistant winter

wheat. The findings from this study, together the studies of Gervais et al (2003) and

Paillard et al (2004), insinuate the involvement of different genomic regions for FHB

resistance in winter wheat compared to spring wheat cultivar Sumai 3 and its

derivatives. Our study also detected a minor FHB QTL on 7AL at NIAB2005. Similar to

our studies, Jia et al (2005), Mardi et al (2006), Semagn et al (2007) and Zhang and

Mergoum (2007) also reported a minor QTL on 7AL in Spring wheats Wangshuibai,

Frontana, NK93604 and Stoa, respectively. In the current study, when FHB resistance

was measured in terms of incidence (Type I- resistance to initial infection), severity

(Type II –resistance to the spread of fungus within the spikelet) and disease index

(Type I and Type II) in C2005 and used in QTL analysis, it was clear that incidence

and severity may be useful to identify the QTL specifically expressed for Type I and

Type II resistance, respectively while disease index could be useful to detect QTL for

field resistance.

In summary, five FHB resistant QTL were identified in this study and two of

these QTL co-localized with semi-dwarfing alleles, Rht-D1b and Rht-B1b. A series of

experiments on the isogenic lines showed that compared to Rht-D1b, Rht-B1b has no

negative effect on resistance under field conditions (low disease pressure) w. Further,

Rht-B1b has a positive effect with regards to Type II resistance compared to Rht-D1b

allele. It is therefore, Rht-B1b is the better dwarfing allele to use in breeding

programmes as it has no detrimental effects on FHB resistance under field conditions.

4.4.3 FHB resistance QTL detected in RL4137/Timgalen cross

4.4.3.1 Disease assessment and QTL mapping for FHB resistance

RL4137 is Canadian awnless, red-grained spring wheat which is highly resistant to

FHB. It is known to have derived from a Brazilian cultivar Frontana, a widely used FHB

resistant source in breeding programmes after Sumai 3. The classical genetic studies

61

have shown the presence of a minimum of two or three additive genes in Frontana

(Singh et al, 1995; VanGinkel et al, 1996). Timgalen is white-grained Australian

spring wheat and is moderately resistant to FHB. It carries an introgressed region on

2B from Triticum timopheevi (AAGG) (Devos et al, 1993). RILs of a cross between

RL4137 and Timgalen were screened for FHB resistance in two polytunnel and a field

experiment using Fu42, a DON-producing isolate of F. grameniarum. Generally,

distribution of AUDPC was continuous and was marginally skewed towards RL4137.

Few transgressive segregants towards RL4137 were observed. The correlation for

AUDPC between experiments was highly significant. Similar to AUDPC, FHB related

traits PH, awns and SW also showed continuous distributions. PH and SW showed a

strong significant negative relationship with AUDPC. The linkage map was constructed

using 341 loci majority of which were DArT (236) and AFLP (90) which organized into

20 major and 24 minor groups. The map based QTL analysis detected consistent QTL

on 1B (Qfhs.jic-1b), 2B (Qfhs.jic-2b) and 6B (Qfhs.jic-6b). The study also indentified

three QTL for FHB resistance which were detected only once either in the polytunnel

or field experiment. The alleles from RL4137 contributed for most of the QTL identified

in this study. The study also identified a positive effect QTL on 6B coming from

Timgalen which is moderately resistant to FHB. It is not uncommon that a moderately

resistant or susceptible parent contributing alleles for resistance when used as one of

the parent with a resistant line. The transgressive segregation of FHB resistance

occurs when wheat with varying levels resistance to FHB are crossed. For example,

Waldron et al (1999) reported a QTL for FHB resistance from a moderately susceptible

parent Stoa, similarly Alondra, a FHB susceptible parent contributed the favourable

allele (Shen et al, 2003).

Poor representation of the D genome in the map and non-representation of 5A

chromosomes in the map may have had some effect on our ability to identify QTL for

FHB resistance and associated traits. Further, majority of the makers used in the

study were DArTs and AFLPs and thus, we were only able to assign the LG to wheat

chromosomes and further extrapolation of QTL positions on the chromosome arms

was not possible.

4.4.3.2 QTL detection for other traits and their role in FHB resistance

Studying the relationship between FHB resistance and other important agronomic

traits is crucial for the development of resistant cultivars. The breeding programmes

can be seriously affected if FHB resistance is linked to undesirable traits. However,

some of the morphological traits associated with FHB resistance potentially limit the

62

number of spores reaching the infection sites on heads and/minimise the chances of

spores gaining entry into head tissues (Steiner et al, 2004). Often association of traits

such as PH, awns, spikelet number, flowering time and other agronomic traits with

FHB resistance have been studied by several workers (Ban and Suenaga, 2000;

Buerstmayr et al, 2000; Draeger et al, 2007; Hilton et al, 1999; Jiang et al, 2006; Liu

et al, 2007; Mesterhazy, 1995; Srinivasachary et al, 2008). In the current study, the

possible role of PH, awns and SW with FHB resistance were investigated.

The PH has been considered as an escape mechanism. Generally, it has been

shown that taller plants alter the micro-climate by having heads further from the soil

creating less favourable conditions for FHB infection and symptom development (Klahr

et al, 2007; Somers et al, 2003; Hilton et al, 1999; Miedaner, 1997; Steiner et al,

2004). In the current study, PH showed significant negative correlation with AUDPC

which could be due to more favourable microclimate for Fusarium infection in short

genotypes. QTL mapping detected a major PH QTL which co-localized with a major

FHB resistant QTL on 2B and the alleles from RL4137 contributed a positive effect for

both. This is in agreement with co-incident QTL for FHB resistance and PH reported in

several studies (Gervais et al, 2003; Paillard et al, 2004; Somers et al, 2003; Steiner

et al, 2004). However, some of the studies have also revealed a more complex

relationship between FHB resistance and PH (Draeger et al, 2007; Gervais et al, 2003;

Paillard et al, 2004; Schmolke et al, 2005; Somers et al, 2003; Srinivasachary et al,

2008). These studies suggested that co-incidence of QTL for FHB resistance and PH

has a genetic basis, linkage or pleiotropy, rather than being due to escape. The study

also identified two minor PH QTL which did not co-localize with any of the FHB

resistance QTL identified in this population. While PH may be correlated with

resistance to FHB, the finding that some QTL for PH coincide with those for FHB, while

others do not, supports this view.

The precise role of awns with FHB resistance in not clear and the literature is

often contradictory. Tamburyic-llincic et al (2007) showed that awned genotypes

showed lower FHB index than awnless genotypes and similarly Ban and Suenaga

(2000) reported that fully awned genotypes were more resistant than tip-awned

genotypes. Further, Snijiders (1990) reported that the association of awns with FHB

are genetically linked and suggested that awns could be used as a marker to select

the resistant lines in the progenies if the resistant parent involved carry awns.

However, Buerstmayr et al (2002) mapped a QTL for FHB resistance on 5A in a Sumai

3 derived line CM-82036 that was not associated with the presence or absence of

awns. Similarly, the studies on Ernie by Liu et al (2007) showed that the presence of

63

awns was not associated with FHB resistance. In the current study, a major QTL for

awns contributed by the alleles from Timgalen on 2B co-localized with a major FHB

resistant QTL contributed by the alleles from RL4137. Most of the variation in the

population for SW was explained by two QTL (2B and 6A) contributed by the alleles

from RL4137 and one of which on 2B co-localized with a major FHB QTL. Co-

localization of QTL for FHB resistance and spike architecture have been reported eg. in

wheat (Draeger et al, 2007) and barley (Ma et al, 2000; Zhu et al, 1999).

As a measure of FHB resistance/tolerance to DON toxin, a trichothecene

produced by Fusarium species was tested on the germination and the growth of wheat

seeds as described in Lemmarks et al (1994) and Gosman et al (2005). The degree of

germination retardation was expressed as the area under disease progress curve

(AUGRC) calculated from seven measurements over ten days. The AUGRC of seed

exposed to DON was expressed as a percentage of the AUGRC of controls. QTL

analysis identified two minor insignificant QTL one each on 2B and 7A with the closest

markers S24/M16I and wPt-6273, respectively which co-localized with the FHB QTL

and both the QTL were contributed by the alleles from RL4137. Moderate correlation

between visual disease scores, DON accumulation in the kernels and in vitro DON

tolerance was reported by Lemmens et al (1994; 1997). Further, the studies by Wang

and Miller (1988) reported that FHB resistant cultivars were more tolerant to DON

than those of FHB susceptible cultivars, however, the experiments by Bruins et al

(1993) and Snijders (1990b) provided an evidence for the lack of link between FHB

resistance and DON tolerance.

The current study identified two major consistent QTL one each on 2B and 6A,

contributed by the alleles from RL4137 and Timgalen, respectively. The QTL for the

FHB related traits PH, SW and awns co-localized with the FHB resistant QTL on 2B.

Co-localization of QTL may be due to linkage or pleiotropy and it has been

hypothesized that the developmental architecture traits determine QTL for FHB

resistance (Zhu et al, 1999). At the level of resolution afforded by this mapping

population, whether the co-localization is due to linkage or pleiotropy could not be

distinguished. Therefore, more studies are necessary to determine whether it is due to

linkage or pleiotropy before attempting to introgress resistance alleles. However, 2B

chromosome of Timgalen carries a large DNA segment from T. timopheevi (Devos et

al, 1993) which makes it less amenable for recombination in that region for precise

mapping of QTL in this region. Hitherto, another mapping population involving RL4137

as one of the parent might be ideal to define the precise position of FHB QTL on 2B.

64

The study provides a starting point for manipulating RL4137 derived resistance in

wheat.

65

5 Identification of new sources of resistance from CIMMYT wheat lines

5.1 Background

Fusarium head blight (FHB) is a destructive disease of wheat worldwide (Parry et al,

1995; Waldron et al, 1999). The predominant causal agents Fusarium graminearum

and F. culmorum, both reduce yield and can contaminate grain with mycotoxins that

render it unsuitable for human and livestock consumption (Gilbert and Tekauz, 2000).

Thus, FHB can be major threat for producers, processors and consumers of wheat.

Agronomic practices and fungicides are not fully effective in controlling the disease

and thus breeding of resistant cultivars has been the strategy adopted to minimize

yield and grain quality losses (Shen et al, 2004).

Resistance of wheat to FHB appears to be horizontal and non-species specific

with no clear evidence for host by pathogen species interaction (van Eeuwijk et al,

1995). Several components of resistance to FHB have been proposed, of which two

have been commonly accepted, Type 1 and Type 2 (Schroeder and Christensen 1963).

Resistance to initial infection (Type 1) is assessed as disease incidence following

natural infection or inoculation by spraying heads at mid-anthesis with conidia

(Miedaner et al, 2003). Resistance to spread within the head (Type 2) is assessed by

injection of inoculum into single florets within the head.

Use of point and spray inoculation in conjunction with molecular mapping has

identified several major quantitative trait loci (QTL) conditioning predominantly Type

II resistance (Anderson et al, 2001; Buestmayr et al, 2002; Shen et al, 2003a, 2003b)

but only a few studies have identified QTL for Type I resistance (Buerstmayr et al,

2002; Steiner et al, 2004; Steed et al, 2005). This may reflect a paucity of Type I

resistance in the germplasm under study, but it is also probable that the need to infer

Type I resistance is hampering the identification of this form of resistance. If species,

or isolates, that produce little or no toxin can infect but not spread within the spike

they might be used as tools to identify Type I resistance. The aim of the current study

was to 1) indentify a method for reliable and easy identification of Type 1 resistance

2) identify and characterise potential sources of FHB resistance among a collection of

winter wheat lines obtained from the CIMMYT, Mexico.

66

5.2 Results

5.2.1 Experiment 1: Characterisation of FHB resistance in wheat

varieties inoculated with toxin producing and non-producing species in

the glasshouse

Thirty Europen winter wheat varieties were inoculated in the glasshouse with either a

single DON-producing isolate of F. culmorum or an isolate of the toxin non-producing

species M. majus and assessed for resistance to FHB. Spray inoculation was used to

estimate the level of combined Type I plus Type II resistance and point inoculation

was used to detect Type II resistance alone. Point inoculation experiment was

repeated in 2005.

5.2.1.1 Spray inoculation

Although the average disease severity caused by F. culmorum (51.5%) was

significantly (P < 0.001) higher than that caused by M. majus (14.8%) the coefficient

of correlation for the relationship between pathogens was moderate (0.52, P =

0.003). For inoculations with F. culmorum the most resistant variety was Spark

(31.79% (The low score for Orvantis (25.67%) was due to late inoculation of this

variety), and the most susceptible was Riband (79.4%), however for inoculations with

M. majus, the most resistant variety was Grief (4.39%) and the most susceptible was

Riband (30.88%) (Table 5.1).

5.2.1.2 Point inoculation

Average disease severity (number of damaged spikelets per head) caused by F.

culmorum in 2004 ranged from 12.17 (Goodwood) to 0.70 (Dekan) with an average of

6.74. M. majus was, however, generally unable to spread beyond the spikelet into

which it was inoculated (Table 5.1). F. culmorum produced similar levels of disease

when the point experiment was repeated in 2005, ranging from 11.22 (Riband) to

1.54 (Claire) with an average of 5.80, but M. majus was, once again, unable to spread

beyond the inoculated spikelet (Table 3). Coefficients of correlation for the relationship

between spray and point inoculation with F. culmorum in 2004 were moderate (0.48,

P = 0.006), but low between point inoculations over years (0.37, P = 0.041).

However, for M. majus, the relationship between point and spray inoculation in 2004

and between point inoculations over years was non-significant (P > 0.05).

67

Table 5.1 Average visual disease scores under controlled environment conditions (glasshouse) for a collection of winter wheat varieties after spray (21 days post inoculation (dpi) and point inoculation (14 dpi) with a deoxynivalenol (DON) producing isolate of F. culmorum and the toxin non-producing species M. majus in (2004) and (2005) Variety

Spray 2004 Point 2004 Point 2005 Spray 2004 Point 2004 Point 2005A43-02 56.36 8 5.6 11.51 0.91 0.25Batis 38.15 4.04 4.96 13.36 1.06 0.35Bentley 51.12 5.3 10.58 16.98 0.89 0.32Biscay 54.41 7.54 4.42 15.53 0.84 0.46Centrum 48.04 5.09 5.36 11.69 0.86 0.13Charger 63.97 5.75 8.38 20.49 1.06 0.87Claire 40.07 7.75 1.54 11.5 0.86 1.17Consort 76.51 7.96 2.46 26.02 0.96 1.18Dekan 50.22 0.7 5.34 5.28 0.8 0.5Einstein 59.87 9.15 9.41 14.62 1.03 0.48Goodwoo 45.73 12.17 8.49 15.28 0.95 0.41Grief 45.23 4.67 6.58 4.39 0.74 0.07Istabraq 41.8 5.88 3.32 10.27 0.96 0.36Napier 46.32 4.86 3.83 11.11 0.8 0.84Nirvana 46.45 3.52 4.2 4.61 1.17 0.49Nijinsky 42.79 5.67 1.89 12.01 1.01 1.01Orvantis 25.67 8.39 7.42 11.27 1.26 0.68Quest 38.08 4.1 7.16 9.65 0.93 0.67Renan 48.04 7.1 2.42 6.4 1.05 0.28Riband 79.4 9.83 11.22 30.88 0.86 1.32Richmond 43.43 3.98 4.5 11 1.05 0.53Robigus 55.43 9.67 4.25 16.72 1.03 0.83Savannah 59.9 7.81 9.16 24.55 0.96 0.32Scorpion 78.74 10.19 9.45 26.75 1.13 0.4Smuggler 39.39 2.03 5.08 12.08 0.86 0.26Spark 31.79 2.9 1.75 4.56 1.11 0.67Tanker 64.76 6.2 9.26 16.73 0.95 0.86Warlock 74.29 11.08 7.37 19.76 1 0.4Winnetou 40.79 6.23 5.42 7.22 0.99 0.23Wizard 46.59 10.01 8.76 21.97 1.01 0.93

F. culmorum M. majus

5.2.2 Characterisation of resistance

The contrasting spreading capabilities of F. culmorum (able to spread) and M. majus

(unable to spread) and method of inoculation were used to infer the presence of Type

1 and/or Type II resistance in the wheat varieties according to their reaction to each

pathogen, e.g. M. majus spray (Type I only), F. culmorum point (Type II only) and F.

culmorum spray (Type I + II) (Table 5.1). For example, Spark was shown to possess

both Type I and Type II resistance as revealed by its response to M. majus spray

(4.56) and F. culmorum point (2.3 (mean over 2004 and 2005) inoculation. It had a

high level of overall resistance (Type I+II), being resistant (31.79) to F. culmorum

spray. Similarly, Dekan was resistant to M. majus spray (5.28) and F. culmorum point

68

(3.1) inoculation indicating that it possessed both Type I and Type II resistance.

However, its overall resistance was only moderate as shown by its response to F.

culmorum spray inoculation (50.22). Grief (5.6) and Renan (4.8) appeared less

resistant to the F. culmorum point inoculation than to the M. majus spray inoculation

(4.39 and 6.4) and were assumed to possess predominantly Type I resistance.

Varieties like Batis (4.5), Napier (4.3) and Istabraq (4.6) were relatively more

resistant to F. culmorum point inoculation than to M. majus spray (13.36, 11.11 and

10.27) inoculation and were assumed to possess predominantly Type II resistance.

The combined resistance of these varieties was however, only moderate (F. culmorum

spray) with the exception of Batis (38.15).

5.2.3. Spray and point inoculation of Mercia winter wheat with DON

and NIV producing isolates of F. graminearum

In 2005, Mercia variety was spray or point inoculated with either a DON (UK1) or NIV

(F86) producing isolate of F. graminearum in two separate experiments with different

spore titres. Over all, both in low (5 ml per spike of 1 x 104 ml-1 for spray and 10 μl of

1 x 105 ml-1 for point) and high titre (5 ml per spike of 1 x 105 ml-1 for spray and 10 μl

of 1 x 106 ml-1 for point) experiments, it was evident that both DON and NIV-

producing isolates produced similar disease levels in point and spray experiments (Fig.

5.1). In the spray inoculation experiment, the DON and NIV isolates were able to

infect and cause similar disease levels. In the point inoculation experiment, the DON-

producing isolate was able to infect and spread rapidly to other spikelets within the

ear, contrastingly, the NIV-producing isolate was largely restricted to the inoculated

spikelet until 18dpi in both low and high titre experiments indicating the inability of

NIV-producing isolate to spread beyond the point of inoculation.

69

Fig. 5.1 Predicted mean scores on Mercia following spray or point inoculation with F. graminearum of different chemotypes UK1 (DON) and F86 (NIV). LT and HT refer to low and high titres used for inoculation

5.2.4 Spray and point inoculation of CIMMYT wheat lines with NIV and

DON-producing isolates of F. graminearum

A total of 300 genotypes developed in the FHB breeding programme at International

Maize and Wheat Improvement Center (CIMMYT), Mexico were screened both in field

and polytunel for to identify Type 1 and Type II resistance and characterise them.

These genotypes were developed at CIMMYT, Mexico and were kindly provided by Dr.

Maarten van Ginkel. Initially all the 300 genotypes were screened for field resistance

by spraying DON-producing isolate of F. graminearum and then 90 best lines were

further characterised. The range of AUDPC scores across the 300 lines assessed in the

field trial of 2005 was very high, ranging from 41 to 2110 with a mean of 412. In

2006, the disease levels (AUDPC) induced by the two chemotypes of F. graminearum

were ranged from 11 to 579 for the DON-producing isolate and 21-469 for the NIV

producing isolate and disease levels were significantly higher for the DON-producing

isolate at all three score dates (Table 5.2). The lines also differed significantly in

resistance at 21, 28 and 32 dpi (Table 5.2). The interaction due to DON and NIV

producing isolates was not significant at 21 dpi (P=0.065) but was highly significant

0

10

20

30

40

50

60

70

80

LT-12dpi LT-18dpi LT-23dpi LT-12dpi LT-18dpi LT-23dpi

Pre

dict

ed m

ean

scor

es

0

10

20

30

40

50

60

70

80

90

HT-12dpi HT-18dpi HT-23dpi HT-12dpi HT-18dpi HT-23dpi

Pre

dict

ed m

ean

scor

es

0

2

4

6

8

10

12

14

LT-12dpi LT-18dpi LT-23dpi LT-12dpi LT-18dpi LT-23dpi

Pre

dict

ed m

ean

scor

es

DON NIV

Point inoculation

DON NIV

Spray inoculation

0

2

4

6

8

10

12

14

16

HT-12dpi HT-18dpi HT-23dpi HT-12dpi HT-18dpi HT-23dpi

Pre

dict

ed m

ean

scor

es

DON NIV

Point inoculation

DON NIV

Spray inoculation

0

10

20

30

40

50

60

70

80

LT-12dpi LT-18dpi LT-23dpi LT-12dpi LT-18dpi LT-23dpi

Pre

dict

ed m

ean

scor

es

0

10

20

30

40

50

60

70

80

LT-12dpi LT-18dpi LT-23dpi LT-12dpi LT-18dpi LT-23dpi

Pre

dict

ed m

ean

scor

es

0

10

20

30

40

50

60

70

80

90

HT-12dpi HT-18dpi HT-23dpi HT-12dpi HT-18dpi HT-23dpi

Pre

dict

ed m

ean

scor

es

0

10

20

30

40

50

60

70

80

90

HT-12dpi HT-18dpi HT-23dpi HT-12dpi HT-18dpi HT-23dpi

Pre

dict

ed m

ean

scor

es

0

2

4

6

8

10

12

14

LT-12dpi LT-18dpi LT-23dpi LT-12dpi LT-18dpi LT-23dpi

Pre

dict

ed m

ean

scor

es

DON NIV

Point inoculation

0

2

4

6

8

10

12

14

LT-12dpi LT-18dpi LT-23dpi LT-12dpi LT-18dpi LT-23dpi

Pre

dict

ed m

ean

scor

es

DON NIV

Point inoculationDON NIV

Point inoculation

DON NIV

Spray inoculationDON NIV

Spray inoculation

0

2

4

6

8

10

12

14

16

HT-12dpi HT-18dpi HT-23dpi HT-12dpi HT-18dpi HT-23dpi

Pre

dict

ed m

ean

scor

es

0

2

4

6

8

10

12

14

16

HT-12dpi HT-18dpi HT-23dpi HT-12dpi HT-18dpi HT-23dpi0

2

4

6

8

10

12

14

16

HT-12dpi HT-18dpi HT-23dpi HT-12dpi HT-18dpi HT-23dpi

Pre

dict

ed m

ean

scor

es

DON NIV

Point inoculationDON NIV

Point inoculation

DON NIV

Spray inoculationDON NIV

Spray inoculation

70

by 28 dpi (P=0.002) became non-significant by 32 dpi (P=0.179) reflecting on the

relative impact of different resistance components at each score date.

71

Table 5.2 ANOVA for disease scores on CIMMYT lines following spray inoculation with a DON (UK1) or NIV (F86)

producing isolate for F. graminearum in the 2006 field trial.

Disease scores 21DPI 28DPI 32DPI AUDPC Source of variation DF MS VR F pr. MS. VR F pr. MS VR F pr. MS VR. F pr.

Pathogen 1 679.5 51.7 <.001 2422.1 67.3 <.001 3073.8 31.66 <.001 355234 46.87 <.001

Line 59 86.8 6.6 <.001 461.1 12.8 <.001 839.7 8.65 <.001 88088 11.62 <.001

Pathogen*Line 59 17.6 1.3 0.065 62.5 1.7 0.002 116.1 1.2 0.179 9554 1.26 0.118

Residual 230 13.1 36 97.1 7579

Total 349 28.2 119.2 234.4 22519

Repeatability 0.83 0.88 0.88 0.9

DF = degrees of freedom, MS = mean square, VR = variance ratio, F pr. = F probability, DPI = days post inoculation.

72

In 2006 field experiment with spary inoculation with a DON producing isolate, genotype

54, 91, 105, 112 and 146 expressed high level of FHB resistance (Table 5.3). The

genotypes 54, 105 and 146 also showed high level of resistance to the NIV isolate

indicating that they also possess a high level of Type I resistance. Other lines, most

strikingly 128 exhibited a high level of resistance to the NIV producer while appearing less

resistant to the DON producer indicating that they their FHB resistance is predominantly

of Type I. Many lines exhibited a high level of Type II resistance following point

inoculation. Of these, only 105 and 112 had also expressed high overall resistance

(product of Type I and Type II) following spray inoculation with the DON isolate in the

field trials. The study identified that line 54, 210 and 232 seem to possess high levels of

Type I resistance.

73

Table 5.3 List of potential CIMMYT genotypes for FHB resistance identified using DON or

NIV producing isolates of F. graminearum or F. culmorum

Wheat F. c. DON (Fu42) Spray

F.g. DON (UK1) F. g. NIV (F86) F. g. DON (UK1)

Line/variety Spray Spray Point

J05-AUDPC J-06 AUDPC

Type I+II

AUDPC Type I AUDPC Type II

7 133.5 579.2 S 271.7 S 62.9 S

14 247 530 S 229 MR 62.3 S

18 222 249.7 MR 111.3 MR 41 MR

24 192 424.2 S 112.7 MR - -

33 151 190.8 MR 74.7 MR 33.1 MR

54 119 24.3 R 14.3 R 50 S

55 109 94.3 MR 23 R 48.3 MR

61 158 187 MR 23.3 R 31.2 MR

62 172 202.3 MR 44 MR 90.8 S

66 61 75.2 R 21 R 38.1 MR

74 174.5 386 MR 67 MR 44.9 MR

80 216 172.8 MR 68 MR 44 MR

83 109 58.7 R 41 MR 36.4 MR

87 133.5 313.2 MR 81.7 MR 88.1 S

88 192 265.8 MR 65.5 MR - -

90 133.5 57.7 R 22 R 40.5 MR

91 130 48 R 28.7 R 42 MR

95 161.5 558.8 S 181.7 MR 51.6 S

103 81 71.5 R 29 R 11.9 R

105 116.5 22.5 R 12.3 R 21.3 R

108 85 92.5 R 20.3 R 13.8 R

110 65 114.5 R 20.3 R 13.9 R

112 120.5 38.8 R 22 R 19.3 R

116 155 259.8 MR 81.7 MR 67.5 S

122 192.5 53 R 15.3 R 43.5 MR

128 102 118.3 R 12.7 R 39.2 MR

132 126 132.7 MR 54 MR 66 S

146 186.5 10.8 R 12.3 R 48.5 MR

147 132.5 203.7 R 43.3 MR 22.2 R

161 198.5 146.7 MR 77.7 MR 43.5 MR

168 89 160.8 MR 60 MR 71.4 S

171 212 521.7 S 159 MR 92.2 S

172 219 307.8 MR 158.3 MR 26.9 MR

173 105.5 62.7 R 41.7 MR 20.6 R

174 218 278.3 MR 86.7 MR 33 MR

]

74

Table 5.3 continued List of potential CIMMYT genotypes for FHB resistance identified using DON or NIV producing isolates of F. graminearum or F. culmorum Wheat F. c. DON (Fu42)

Spray F.g. DON (UK1) F. g. NIV (F86) F. g. DON (UK1)

Line/variety Spray Spray Point

J05-AUDPC J-06 AUDPC

Type I+II

AUDPC Type I AUDPC Type II

176 124 340 MR 50 MR 30.9 MR

178 133 194 MR 59.7 MR 88.6 S

182 123 383.3 MR 130.7 MR 65.5 S

183 187 179.2 MR 81.7 MR 56.1 S

186 71.5 55.7 R 14.3 R 29.1 MR

187 170.5 57.5 R 18.3 R 19 R

189 193 418.4 S 63.3 MR 44.9 MR

193 109 191.5 R 37 R 36.4 MR

194 196 79.5 R 25 R 23.4 R

195 138 67.2 R 18.3 R 40.9 MR

200 237 372.8 MR 185.7 MR 38.3 MR

205 143 71.8 R 28.3 R 23.4 R

206 105 400 S 144 MR 30.1 MR

209 109 227.3 MR 44.3 MR 28.4 MR

210 103 96.5 R 29.7 R 83.8 S

224 165 233.2 MR 82 MR 90.4 S

232 55 59.5 R 34 R 62.4 S

233 122 177.7 R 29 R 40.9 MR

234 259 296.6 MR 58 MR 39.5 MR

240 161.5 147.2 R 30 R 39.8 MR

241 162 425 S 144 MR 68.7 S

244 274 152.5 MR 82.7 MR 38.8 MR

251 95.5 53 R 13.3 R 19.4 R

257 200 298.8 MR 53 MR 61.3 S

264 202 485 S 96.5 MR - -

268 194.5 337.5 MR 78.7 MR - -

Renan - 75.6 R 17.8 R - -

Spark - 93.2 R 39.4 R - -

Riband - 632.2 S 265.4 S - -

Sumai 3 - - - - - 7.7 R

Timgalen - - - - - 43.3 MR

Remus - - - - - 70 S

F. c. = Fusarium culmorum, F. g. = F. graminearum, J05 = JIC2005, J06 = JIC2006, JIC = JIC2007, AUDPC = area under disease progress curve; R, resistant; I, moderately resistant; S, susceptible; -, no data Type I and Type II refers to resistance to initial infection and resistance to spread with in the spike (sensu Schroeder & Christensen (1963).

75

5.3 Discussion

The resistance of wheat to FHB is of two types: resistance to initial infection (now termed

type I) and resistance to spread within the plant (now termed type II) (Schroeder and

Christensen 1966) and the varieties may contain either or both types of resistance.

However, much research has concentrated upon identifying and mapping type II

resistance (see Lin et al, 2006). Generally type I resistance evaluation is more tricky and

is carried out in field trials using spray or natural inoculation and measuring disease as the

percentage of infected heads in plots or as disease severity (Steiner et al, 2004; Lin et al,

2006). The DON myco-toxin acts as a virulence factor required to enable spread of F.

graminearum between spikelets in earheads (Jansen et al, 2005). The isolates of F.

grameniarum with disrupted in Tri5, a gene essential for trichothecene biosynthesis (non-

DON producers), are unable to spread from inoculated floret (Proctor et al, 1995). The

current study species, M. majus or a NIV-producing isolate of F. graminearum could be

used to screen for resistance to initial infection (type I resistance) minimising the

confounding effects caused by spread within the wheat head.

Following spray inoculation, both F. culmorum and M. majus produced disease

although M. majus was much less aggressive under the conditions used herein. Numerous

reports suggest that M. majus is a weak pathogen of wheat heads (Brennan et al, 2005).

In contrast, following point inoculation, M. majus was unable to spread beyond the

infected spikelet whereas the DON-producing F. culmorum spread into the rachis and

throughout the head. The few studies that have compared the resistance of wheat

varieties to different FHB species concluded that resistance acts similarly against all

species (Mesterhazy et al, 2005; Toth et al, 2008). However, no comprehensive

comparison has been made for M. majus and Fusarium species. It has been reported that

the infection processes and colonisation of wheat florets by M. majus is similar to that of

F. graminearum and F. culmorum but spread into adjacent spikelets was not noted (Kang

et al, 2004). Symptoms produced by M. majus, a non toxin-producing species (Jennings,

2005) are almost identical to those produced by Tri5-transformants (Cuzick et al, 2008),

being restricted to single spikelets and unable to spread throughout the head. Our suggest

that spray inoculation with M. majus may be used to assess Type I resistance directly in

the absence of any confounding effects caused by differing levels of Type II resistance.

This technique would complement point inoculation with a DON-producing isolate to

evaluate Type II resistance. Results following spray inoculation with the DON-producing

76

isolate of F. culmorum were deduced to represent the product of Type I and Type II

resistance in each variety.

A number of the varieties examined appeared to lack appreciable levels of either

Type I or Type II resistance. These included Riband, Charger, Scorpion 25, Tanker,

Consort and Goodwood. In many cases, combination of moderate and resistant Type I and

Type II components appeared to be additive and resulted in a higher level of overall

resistance. The variety Spark exhibited the best level of overall resistance and appeared

to possess both Type I and Type II resistance. The variety Grief had the most contrasting

response to point inoculation with F. culmorum and spray inoculation with M. majus with

the relatively high level of resistance against the latter indicating that it possesses

predominantly Type I resistance with very little Type II. The varieties Nirvana, Centrum

and Renan also appeared to possess greater Type I than Type II resistance but the effect

was less marked than that of Greif. However, there were also many cases where

combinations of moderate or resistant components did not appear to be additive.

Combining ability analysis in wheat has identified both general (GCA) and specific (SCA)

combining effects associated with FHB resistance (Buerstmayr et al, 1999; Mardi et al,

2004). In addition, genetic analysis of combinations of quantitative trait loci (QTL) by

Miedaner et al (2006) indicated that certain combinations of QTL conditioning Type I and

Type II components produced higher levels of resistance than others hinting at the

presence of additive by additive epistatic effects.

Although the non toxin-producing M. majus did infect wheat heads, disease levels

were consistently low under the conditions used herein. Therefore, a more virulent

alternative pathogen was sought to assess Type I resistance. NIV has been shown to be

much less phytotoxic than DON on both wheat and Arabidopsis (Eudes et al, 2000;

Desjardins et al, 2007). While NIV-producing isolates of F. graminearum can spread within

wheat heads, disease progress is often slow (Maier et al, 2006), although this may be due

to a reduced level of toxin biosynthesis rather than to the type of trichothecene (Goswami

& Kistler, 2005). Disease severity on Mercia 12 dpi following spray inoculation with a NIV-

producing isolate (F86) was similar to that induced by a DON producing strain (UK1)

indicating that both strains were equally aggressive during initial infection. In contrast,

following point inoculation the DON producer rapidly spread through the spike, whilst the

NIV isolate remained largely confined to the inoculated spikelet for at least the first 18

days and had only significantly moved beyond the infection site by 23 dpi. Taken

together, the spray and point data suggest that both isolates are equally virulent, but the

77

DON producer is able to move beyond the point(s) of infection to cause greater over-all

disease severity. These data are in accordance with field studies of disease severity in

wheat and rye which reported that NIV producing isolates of F. culmorum caused less

disease and accumulated less toxin in grain samples than DON producing isolates

(Miedaner and Reinbrecht, 2001). While the isolates UK1 and F86 are DON (15ADON) and

NIV chemotypes, respectively and no attempt was made to determine the amounts of

toxin made by them. It is possible that F86 produced only small amounts of NIV in planta

that may account for it not spreading within the spike until very late in infection relative to

the DON producer (UK1). However, as NIV is intrinsically less phytotoxic than other

trichothecenes such as DON (Eudes et al, 1997) and this may explain the reduced ability

of the NIV chemotype to spread within the spike. Whatever the underlying cause, we

conclude that virulent NIV chemotype isolates of F. graminearum, such as F86, might be

used in spray inoculation trials to determine relative levels of Type I resistance in wheat.

We used a combination of point and spray trials with appropriate NIV and DON producing

isolates of F. graminearum to characterise the FHB resistance within 60 lines of wheat

obtained from the wheat breeding of the International Maize and Wheat Improvement

Center (CIMMYT), Mexico. The field trials confirmed the differential rate of spread of the

NIV and DON producing isolates and showed those with high overall FHB resistance and

those with predominantly Type I resistance. Several lines exhibiting high levels of FHB

resistance in field trials were found to also possess high levels of Type II resistance

following point inoculation. More significantly, a few lines (54, 210 and 232), while

exhibiting high levels of FHB resistance in field trials, were highly susceptible in point

inoculation studies indicating that their resistance is predominantly of Type I. While a

relatively large number of Type II resistance genes and QTL have been reported (Shen et

al, 2003; Ma et al, 2006; Cuthbert et al, 2007; 2008), few sources of Type I resistance

have been identified to date (Lin et al, 2006; Steed et al, 2005). This, in large part, is due

to the greater technical challenges associated with the unequivocal identification of Type I

resistance.

Point inoculation with a virulent DON chemotype of F. graminearum or F. culmorum

is recognised as an appropriate method to screen wheat varieties for relative levels of

Type II FHB resistance. On the basis of our results from disease trials, we propose that

spray inoculation with a virulent non DON-producing FHB pathogen may be used to

directly assess relative levels of Type I resistance in the absence of the confounding

effects due to differences in Type II resistance. M. majus may be suitable but this species

78

is generally not highly virulent in our experience. The use of a virulent NIV chemotype

isolate of F. graminearum or F. culmorum would be more appropriate. Some NIV

chemotypes within the F. graminearum species complex have been shown to be able to

spread within wheat spikes (Goswami & Kistler, 2005). Therefore, prior to use in

assessments of Type I resistance, the virulence of NIV chemotype isolates should be

evaluated by both spray and point inoculation to select those (like F86) that are virulent

but unable to spread within the wheat spike. We propose that the use of appropriate non

DON-producing FHB species or isolates in spray inoculation trials combined with point

inoculation using DON-producing isolates will assist researchers seeking to identify and

characterise resistance to FHB in wheat.

6. Evaluation of FHB resistance within UK spring barley varieties

6.1 Background

Relatively little is known about the genetics of FHB resistance in barley, although it is

believed that barley generally has good Type 2 resistance (resistance to spread of

infection within the ear). The implication is that differences in Type 1 resistance

(resistance to initial infection) are likely to be of paramount importance in barley

varieties. There is currently no information available on the resistance of UK barley

varieties to FHB.

The aim of the work described here was to carry out an initial screen of spring

barley varieties on the current UK Recommended List to look for evidence of varietal

differences in resistance to FHB.

6.2 Materials and Methods

Field experiments investigating the resistance of UK spring barley varieties to FHB were

carried out in two years, 2005 and 2006, at NIAB, Cambridge.

2005

27 spring barley varieties, comprising those in current UK Recommended List trials, were

grown in two experiments, each being a randomised block design with 4 replicates. Plots

were approximately 1m x 1m. The first experiment (‘early’) was inoculated with spores of

F. culmorum on 7 June, to coincide with mid anthesis of the earlier flowering varieties.

The second experiment (‘late’) was inoculated three days later on 10 June 2005, to

79

coincide with mid anthesis of the later flowering varieties. A single very late flowering

variety in the late trial was individually inoculated on 14 June. A spore suspension of a

DON-producing isolate of F.culmorum (Fu42, from JIC), at a concentration of 1 x 106

spores / ml, was applied at a rate of 100ml per plot, using a hand held pump sprayer.

Assessment of FHB infection was carried out in the late experiment only, as

disease levels were extremely low in the early experiment. Symptoms were assessed

visually on 4 July, at around GS 80, using an infection index method (% ears infected x

mean severity of infection of infected ears). At harvest, grain samples were taken for

analysis of DON content.

2006

22 spring barley varieties from amongst the 27 tested in 2005 were grown in a single

experiment in 2006. Experimental design was a randomised block with 4 replicates. Plots

were approximately 0.5m x 1m. All plots were inoculated twice, the earlier inoculation,

on 9 June 2006, being when the earliest flowering varieties started to flower and the

later inoculation, on 12 June, being when the majority of other varieties reached

anthesis. A few late flowering varieties required an additional inoculation on 15 June.

Inoculation methods were similar to those in 2005, except that 50ml of inoculum was

applied per plot, to reflect their smaller size.

Symptoms of FHB were assessed on three dates, 22 June (GS 71), 29 June (GS

83) and 7 July (GS 85), using the same method as in 2005.

6.3 Results

FHB infection was considerably more severe in the 2006 trial than the 2005 trial.

However, in both years there were highly significant differences in infection level

between varieties, with a significant correlation between years for the 22 common

varieties (r = 0.857, P<0.001).

Table 6.1 summarises assessments of FHB symptoms in the two years’ trials. Varieties

were ranked in order of severity of infection in 2006. The data for 2005 are from a single

assessment, whereas the 2006 data are means of three assessments made over a period

of 15 days.

80

Table 6.1. FHB severity (0-100) in two years’ trials.

Variety

FHB severity in 2005 trial (0-100)

FHB severity in 2006 trial (0-100)

Cocktail 11.29 51.67 Cellar 8.23 34.08 Doyen 7.08 33.17 NFC Tipple 6.28 27.83 Static 10.52 27.75 Power 3.05 24.83 Hydra 6.84 23.75 Wicket 4.69 22.08 Kirsty 2.68 21.08 Tocada 0.72 20.58 Decanter 2.49 19.83 Poker 1.95 19.33 Chalice 4.30 17.83 Riviera 2.70 17.50 Rebecca 0.75 17.17 Waggon 4.49 15.84 Oxbridge 2.12 14.51 Optic 0.71 14.51 Appaloosa 1.46 13.34 Spire 0.55 11.68 Troon 0.08 10.68 Westminster 0.92 9.51 Putney 6.32 - Carafe 2.70 - Cribbage 2.16 - Centurion 2.05 - Beatrix 2.03 - Anovar Variety effect P value <0.001 <0.001 SE difference 1.906 3.468 lsd P<0 05 3.775 6.868

81

Table 6.2 compares the results of DON analyses of grain samples taken from the 2005

trial (bulked over replicates) with FHB severity. There was a significant correlation

between the two measures (r = 0.747, P<0.001).

Table 6.2. FHB severity and DON content of grain in 2005

Variety

FHB severity (0-100)

DON (ppm)

Cocktail 10.72 1.50 Static 10.07 1.70 Cellar 7.95 1.52 Doyen 7.16 3.05 Hydra 6.61 2.08 Chalice 4.20 1.10 Wicket 4.00 0.70 Waggon 3.75 1.00 Power 2.63 1.90 Kirsty 2.47 0.30 Decanter 2.31 0.81 Riviera 2.11 0.65 Oxbridge 1.85 0.20 Poker 1.68 0.29 Aquila 1.39 0.31 Westminster 0.88 0.35 Rebecca 0.73 0.26 Tocada 0.73 0.13 Optic 0.72 0.37 Spire 0.54 0.08 Troon 0.05 0.35

6.4 Discussion

The results obtained from the two trials provide clear indications of significant and

consistent differences between barley varieties in resistance to FHB. More detailed work

is needed to dissect the components of resistance and determine whether the apparent

resistance may be influenced by the flowering characteristics of a variety, in particular its

tendency to flower while the ear is still in boot.

82

7 Acknowledgements

We wish to express our since gratitude to the UK Home-Grown Cereals Authority (HGCA)

and the Department of Environment, Food and Rural Affairs (Defra) through a Defra-

LINK project (No. LK0932). All the authors would like to thank collaborating commercial

partners Nickerson Seeds (UK) Ltd., Elsoms Seeds Ltd., Advanta Seeds (UK) Ltd. and

RAGT (UK) Ltd. (collaboration initiated with Monsanto (UK) Ltd., who subsequently

transferred their wheat breeding programme to RAGT for their invaluable contributions.

We also acknowledge help rendered by all those involved in field and polytunnel related

work.

83

8 References

Anderson JA, Stack RW, Liu S, Waldron BL, Fjeld AD, Coyne C, Moreno-Sevilla B, Fetch JM, Song QJ, Cregan PB, Frohberg RC. 2001. DNA markers for Fusarium head blight resistance QTL in two wheat populations. Theoretical and Applied Genetics, 102:1164-1168

Bai GH, Kolb FL, Shaner G, Domier LL. 1999. Amplified fragment length polymorphism

markers linked to a major quantitative trait locus controlling scab resistance in wheat. Phytopathology, 89:343-348

Ban T, Suenaga K. 2000. Genetic analysis of resistance to Fusarium head blight caused

by Fusarium graminearum in Chinese wheat cultivar Sumai 3 and the Japanese cultivar Saikai 165. Euphytica, 113:87-99

Botstein D, White RL, Skolnick M, Davis RW. 1980. Construction of a Genetic linkage

map in man using restriction fragment length polymorphisms. American Journal of Human Genetics, 32:314 – 331

Bourdoncle W, Ohm HW. 2003. Fusarium head blight-resistant wheat line 'Bizel' does

not contain rye chromatin. Plant Breeding, 122:281-282 Buerstmayr H, Legzdina L, Steiner B, Lemmens M. 2004. Variation for resistance to

Fusarium head blight in spring barley. Euphytica, 137:279-290 Buerstmayr H, Lemmens M, Berlakovich S, Ruckenbauer P. 1999. Combining ability

of resistance to head blight caused by Fusarium culmorum (WG Smith) in the F-1 of a seven parent diallele of winter wheat (Triticum aestivum L.). Euphytica, 110:199-206

Buerstmayr H, Lemmens M, Grausgruber H, Ruckenbauer P. 1996. Scab resistance

of international wheat germplasm. Cereal Research Communications 24:195-202 Buerstmayr H, Lemmens M, Hartl L, Doldi L, Steiner B, Stierschneider M,

Ruckenbauer P. 2002. Molecular mapping of QTL for Fusarium head blight resistance in spring wheat. I. Resistance to fungal spread (type II resistance). Theoretical and Applied Genetics, 104:84-91

Central Science Laboratory. 2004. Fusarium diseases of wheat: Latest press release.

http://www.csl.gov.uk/science/organ/environ/entom/fusarium/ Chandler EA, Simpson DR, Thomsett MA, Nicholson P. 2003. Development of PCR

Assays to Tri7 and Tri13 Trichothecene biosynthetic genes, and characterisation of chemotypes of Fusarium graminearum, Fusarium culmorum and Fusarium cerealis. Physiological and Molecular Plant Pathology, 62:355-367

Cuthbert PA, Somers DJ, Brule-Babel A. 2007. Mapping of Fhb2 on chromosome 6BS:

A cene controlling Fusarium head blight field resistance in bread wheat (Triticum aestivum L.). Theoretical and Applied Genetics, 114:429-437

84

Cuthbert PA, Somers DJ, Thomas J, Cloutier S, Brule-Babel A. 2006. Fine mapping

Fhb1, a major gene controlling Fusarium head blight resistance in bread wheat (Triticum aestivum L.). Theoretical and Applied Genetics, 112:1465-1472

Devos KM, Millan T, Gale MD. 1993. Comparative RFLP maps of the homeologous

group 2 chromosomes of wheat, rye and barley. Theoretical and Applied Genetics, 85:784-792

Draeger R, Gosman N, Steed A, Chandler E, Thomsett M, Srinivasachary,

Schondelmaier J, Buerstmayr H, Lemmens M, Schmolke M, Mesterhazy A, Nicholson P. 2007. Identification of QTL for resistance to Fusarium head blight, DON accumulation and associated traits in the winter wheat variety Arina. Theoretical and Applied Genetics, 115:617-625

Ellis MH, Spielmeyer W, Gale KR, Rebetzke GJ, Richards RA. 2002. “Perfect”

markers for the Rht-B1b and Rht-D1b dwarfing genes in wheat. Theoretical and Applied Genetics, 105:1038-1042

Flintham JE, Borner A, Worland AJ, Gale MD. 1997. Optimising grain yield : Effects of

Rht (gibberllin-insensitive) dwarfing genes. Journal of Agricultural Science Cambridge, 128:11-25

Gervais L, Dedryver F, Morlais JY, Bodusseau V, Negre S, Bilous M, Groos C,

Trottet M. 2003. Mapping of quantitative trait loci for field resistance to Fusarium head blight in an European winter wheat. Theoretical and Applied Genetics, 106:961-970

Gilbert J, Tekauz A. 2000. Review: Recent developments in research on fusarium head

blight of wheat in Canada. Canadian Journal of Plant Pathology, 22:1-8 Gosman N, Bayles R, Jennings P, Kirby J, Nicholson P. 2007. Evaluation and

characterization of resistance to Fusarium head blight caused by Fusarium culmorum in UK winter wheat cultivars. Plant Pathology, 56:264-276

Gosman N, Chandler E, Thomsett M, Draeger R, Nicholson P. 2005. Analysis of the

relationship between parameters of resistance to Fusarium head blight and in vitro tolerance to deoxynivalenol of the winter wheat cultivar WEK0609 (R). European Journal of Plant Pathology, 111:57-66

Guo PG, Bai GH, Shaner GE. 2003. AFLP and STS tagging of a major QTL for Fusarium

head blight resistance in wheat. Theoretical and Applied Genetics, 106:1011-1017 Hilton AJ, Jenkinson P, Hollins TW, Parry DW. 1999. Relationship between cultivar

height and severity of Fusarium ear blight in wheat. Plant Pathology, 48:202-208 Jaccoud D, Peng K, Feinstein D, Kilian A. 2001. Diversity Arrays: a solid state

technology for sequence information independent genotyping. Nucleic Acids Research, 29:e25

85

Jenkinson P, Parry DW. 1994. Splash dispersal of conidia of Fusarium culmorum and Fusarium avenaceum. Mycological Research, 98:506-510

Jia G, Chen PD, Qin GJ, Bai GH, Wang X, Wang SL, Zhou B, Zhang SH, Liu DJ.

2005. QTL for Fusarium head blight response in a wheat DH population of Wangshuibai/Alondra's'. Euphytica, 146:183-191

Jiang GL, Dong YH, Lewis JM, Siler L, Ward RW. 2006. Characterization of resistance

to Fusarium graminearum in a recombinant inbred line population of wheat: Resistance to fungal spread, mycotoxin accumulation, and grain yield loss, and trait relationships. Crop Science, 46:2590-2597

Klahr A, Zimmermann G, Wenzel G, Mohler V. 2007. Effects of environment, disease

progress, plant height and heading date on the detection of QTL for resistance to Fusarium head blight in an European Winter Wheat Cross. Euphytica, 154:17-28

Kosambi DD. 1944. The estimation of Map values from recombination value. Annals of

Eugenics, 12:172-175 Lemanczyk G, Sadowski CK. 2002. Fungal communities and health status of roots of

winter wheat cultivated after oats and oats mixed with other crops. Biocontrol, 47:349-361

Liu S, Abate ZA, McKendry AL. 2005. Inheritance of Fusarium head blight resistance in

the soft red winter wheat Ernie. Theoretical and Applied Genetics, 110:454-461 Ma HX, Zhang KM, Gao L, Bai GH, Chen HG, Cai ZX, Lu WZ. 2006. Quantitative trait

loci for resistance to fusarium head blight and deoxynivalenol accumulation in Wangshuibai wheat under field conditions. Plant Pathology, 55, 739-745

Ma ZP, Steffenson BJ, Prom LK, Lapitan NLV. 2000. Mapping of quantitative trait loci

for Fusarium head blight resistance in barley. Phytopathology, 90:1079-1088 McCartney CA, Somers DJ, Fedak G, Cao W. 2004. Haplotype diversity at Fusarium

head blight resistance QTL in wheat. Theoretical and Applied Genetics, 109:261-271 McCartney CA, Somers DJ, Fedak G, DePauw RM, Thomas J, Fox SL, Humphreys

DG, Lukow O, Savard ME, McCallum BD, Gilbert J, Cao W. 2007. The evaluation of FHB resistance QTL introgressed into elite Canadian spring wheat germplasm. Molecular Breeding, 20:209-221

Mesterhazy A, Bartok T, Mirocha CG, Komoroczy R. 1999. Nature of wheat

resistance to Fusarium head blight and the role of deoxynivalenol for breeding. Plant Breeding, 118:97-110

Mesterhazy A. 1995. Types and components of resistance to Fusarium head blight of

wheat. Plant Breeding, 114:377-386 Miedaner T, Reinbrecht C, Lauber U, Schollenberger M, Geiger HH. 2001. Effects of

genotype and genotype-enviromment interaction on deoxynivalenol accumulation

86

and resistance to Fusarium head blight in rye, triticale, and wheat. Plant Breeding, 120:97-105

Miedaner T, Moldovan A, Ittu M. 2003. Comparison of spray and point inoculation to assess

resistance to fusarium head blight in a multienvironment wheat trial. Phytopathology, 93:1068-1072

Miller JD, Arnison PG, 1986. Degradation of deoxynivalenol by suspension cultures of

the Fusarium head blight resistant wheat cultivar Frontana. Canadian Journal of Plant Pathology, 8:147-150

Nyquist WE. 1991. Estimation of heritability and prediction of selection response in

plant-populations. Critical Reviews in Plant Sciences, 10:235-322 Paillard S, Schnurbusch T, Tiwari R, Messmer M, Winzeler M, Keller B,

Schachermayr G. 2004. QTL Analysis of resistance to Fusarium head blight in Swiss winter wheat (Triticum aestivum L.). Theoretical and Applied Genetics, 109:323-332

Parry DW, Jenkinson P, McLeod L. 1995. Fusarium Ear Blight (Scab) in Small-Grain

Cereals - a Review. Plant Pathology, 44:207-238 Paul PA, Lipps PE, Madden LV, 2005. Relationship between visual estimates of

Fusarium head blight intensity and deoxynivalenol accumulation in harvested wheat grain: A meta-analysis. Phytopathology, 95:1225-1236

Pestka JJ, Smolinski AT. 2005. Deoxynivalenol: Toxicology and potential effects on

humans. Journal of Toxicology and Environmental Health-Part B-Critical Reviews, 8:39-69.

Röder MS, Korzun V, Wendehake K, Plaschke J, Tixier MH, Leroy P, Ganal MW, 1998. A microsatellite map of wheat. Genetics, 149:2007-2023

Schmolke M, Zimmermann G, Buerstmayr H, Schweizer G, Miedaner T, Korzun V,

Ebmeyer E, Hartl L. 2005. Molecular mapping of Fusarium head blight resistance in the winter wheat population Dream/Lynx. Theoretical and Applied Genetics, 111:747-756

Schroeder HW, Christensen JJ. 1963. Factors affecting resistance of wheat to scab

caused by Gibberella zeae. Phytopathology, 53:831-837 Semagn K, Skinnes H, Bjornstad A, Maroy AG, Tarkegne Y. 2007. Quantitative trait

loci controlling Fusarium head blight resistance and low deoxynivalenol content in hexaploid wheat population from 'Arina' and NK93604. Crop Science, 47:294-303

Shen XR, Ittu M, Ohm HW. 2003a. Quantitative trait loci conditioning resistance to

Fusarium head blight in wheat line F201r. Crop Science, 43:850-857 Shen X, Zhou M, Lu W, Ohm H. 2003b. Detection of Fusarium head blight resistance

QTL in a wheat population using bulked segregant analysis. Theoretical and Applied Genetics, 106, 1041-1047

87

Shen R, Kong LR, Ohm H. 2004. Fusarium head blight resistance in hexaploid wheat (Triticum aestivum)-Lophopyrum genetic lines and tagging of the alien chromatin by PCR markers. Theoretical and Applied Genetics, 108:808-813

Simpson DR, Weston GE, Turner JA, Jennings P, Nicholson P. 2001. Differential

control of head blight pathogens of wheat by fungicides and consequences for mycotoxin contamination of grain. European Journal of Plant Pathology, 107:421-431

Singh RP, Ma H, Rajaram S. 1995. Genetic analysis of resistance to scab in spring

wheat cultivar Frontana. Plant Disease, 79:238-240 Snijders CHA. 1990. Genetic variation for resistance to Fusarium head blight in bread

wheat. Euphytica, 50:171-179 Somers DJ, Fedak G, Savard M. 2003. Molecular mapping of novel genes controlling

Fusarium head blight resistance and deoxynivalenol accumulation in spring wheat. Genome, 46:555-564

Somers DJ, Isaac P, Edwards K. 2004. A high density microsatellite consensus map for

bread wheat (Triticum aestivum L.). Theoretical Applied Genetics, 109:1105-1114 Srinivasachary, Gosman N, Steed A, Simmonds J, Leverington-Waite M, Wang Y,

Snape J, Nicholson P. 2008. Susceptibility to Fusarium head blight is associated with the Rht-D1b semi-dwarfing allele in wheat. Theoretical and Applied Genetics, (published online)

Steed A, Chandler E, Thomsett M, Gosman N, Faure S, Nicholson P. 2005.

Identification of type I resistance to Fusarium head blight controlled by a major gene located on chromosome 4A of Triticum macha. Theoretical and Applied Genetics, 111:521-529

Steiner B, Lemmens M, Griesser M, Scholz U, Schondelmaier J, Buerstmayr H.

2004. Molecular mapping of resistance to Fusarium head blight in the spring wheat cultivar Frontana. Theoretical and Applied Genetics, 109:215-224

Tamburic-Ilincic L, Schaafsma AW, Falk DE. 2007. Indirect selection for lower

deoxynivalenol (DON) content in grain in a winter wheat population. Canadian Journal of Plant Science, 87:931-936

Van Eeuwijk, EA, Mesterhazy, A, Kling CI, Ruckenbauer P, Saur L, Buerstmayr H,

Lemmens M, Keizer LCP, Maurin N, Snijders CHA. 1995. Assessing non-specificity of resistance in wheat to head blight caused by inoculation with European strains of Fusarium culmorum, F. graminearum and M. nivale using a multiplicative model for interaction. Theoretical and Applied Genetics, 90:221-228

Van Ooijen JW, Voorips RE. 2001. Joinmap 3.0. Software for the calculation of genetic

linkage maps. Wageningen, the Netherlands: Plant Research International

88

Van Ooijen JW. 2004. MapQTL Version 5.0, Software for the mapping of quantitative trait loci in experimental populations. Kyazma B.V., Wageningen, The Netherlands

Van Sanford D, Anderson J, Campbell K, Costa J, Cregan P, Griffey C, Hayes P,

Ward R. 2001. Discovery and deployment of molecular markers linked to Fusarium head blight resistance: An integrated system for wheat and barley. Crop Science, 41:638-644

VanGinkel M, VanderSchaar W, Yang ZP, Rajaram S. 1996. Inheritance of resistance

to scab in two wheat cultivars from Brazil and China. Plant Disease, 80:863-867 Voorrips RE. 2002. MAPCHART: Software for the graphical presentation of linkage maps

and QTL. Heredity, 93:77-78 Vos P, Hogers R, Bleeker M, Reijans M, Vandelee T, Hornes M, Frijters A, Pot J,

Peleman J, Kuiper M, Zabeau M. 1995. AFLP - a new technique for DNA fingerprinting. Nucleic Acids Research, 23:4407-4414

Waldron BL, Moreno-Sevilla B, Anderson JA, Stack RW, Frohberg RC. 1999. RFLP

mapping of QTL for fusarium head blight resistance in wheat. Crop Science, 39 :805-811

Windels CE. 2000. Economic and social impacts of Fusarium head blight: Changing

farms and rural communities in the Northern Great Plains. Phytopathology, 90, 17-21 Zadoks JC, Chang TT, Konzak CF. 1974. Decimal code for growth stages of cereals.

Weed Research, 14:415-421 Zhou LL, Bai GH, Carver B, Zhang DD. 2007. Identification of new sources of

aluminium resistance in wheat. Plant and Soil, 297:105-118 Zhou WC, Kolb FL, Bai GH, Shaner G, Domier LL. 2002. Genetic analysis of scab

resistance QTL in wheat with microsatellite and AFLP markers. Genome, 45:719-727 Zhu H, Gilchrist L, Hayes P, Kleinhofs A, Kudrna D, Liu Z, Prom L, Steffenson B,

Toojinda T, Vivar H. 1999. Does function follow form? Principal QTL for Fusarium head blight (FHB) resistance are coincident with QTL for inflorescence traits and plant height in a doubled-haploid population of barley. Theoretical and Applied Genetics, 99:1221-1232